HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Pate, J. W. Articles by Walker, W. A. Search for Related Content PubMed PubMed Citation Articles by Pate, J. W. Articles by Walker, W. A.

Ann Thorac Surg 1995;59:90-98
© 1995 The Society of Thoracic Surgeons

Acute Traumatic Rupture of the Aortic Isthmus: Repair With Cardiopulmonary Bypass

Cardiothoracic Surgery Section, College of Medicine, University of Tennessee-Memphis, and Elvis Presley Trauma Center, Regional Medical Center, Memphis, Tennessee

Accepted for publication June 16, 1994.

	Abstract

Top Footnotes Abstract Introduction Material and Methods References

 
In an attempt to prevent paraplegia, a devastating complication common after the repair of traumatic rupture of the aorta, we have used partial cardiopulmonary bypass. Most of the patients in our series (79.5%) underwent other major surgical procedures immediately before or after the aortic repair. Eight of the 110 patients died before aortic repair could be performed. The aorta was not repaired in 3, because of other injuries. In 9, the repair was done without a shunt or bypass; 4 patients died and 2 (22.2%) survived without paraplegia. One of the 2 who underwent repair with a Gott shunt died; the survivor suffered no cord damage. Of the 88 patients whose repair was carried out under cardiopulmonary bypass, 6 died and 80 (90.9%) survived without paraplegia. None of the last 39 patients has become paraplegic, as vasodilator treatment is now discontinued during the cross-clamp period. Serious intracranial injury was present in 19 patients; in 3 (15.8%) the injury became worse after repair. There was no evidence of new or increased intraabdominal bleeding during heparinization. Except in the event of pulmonary lacerations, systemic heparin therapy was not associated with major problems.

	Introduction

Top Footnotes Abstract Introduction Material and Methods References

 
See also page 98.

Ischemic myelopathy of the spinal cord with partial or complete paraplegia is a frequent consequence of the surgical repair of an acute traumatic rupture of the aorta (ATRA). Aortic cross-clamping without the use of some form of bypass is associated with many other deleterious effects: elevated left ventricular pressure (arrhythmia and heart failure), pulmonary edema, right ventricular strain, elevated central venous and cerebrospinal fluid pressures (potentially increasing any intracranial bleeding and decreasing cerebral and spinal blood flow), and ischemia below the clamp (cord ischemia, and renal and hepatic failure).

Decompression of the ascending aorta and heart, and the associated perfusion of the distal body accomplished by shunt or bypass, markedly ameliorate these effects. In an attempt to prevent the paraplegia, we use partial cardiopulmonary bypass (PCPB), and this has been our practice since 1964. Our first 59 patients with ATRA were reported in 1985 [1]; this experience increased to 110 patients by January 1, 1993.

This report addresses two major questions: Can the use of PCPB bring about a decrease in the incidence of paraplegia? and Do the dangers of systemic heparinization exceed the benefits conferred by PCPB?

Although the aorta may be disrupted intrapericardially, in the arch or at distal levels, by far the most common site seen clinically is a ``classic'' disruption of part or all of the circumference of the aorta at the isthmus. We classify traumatic rupture of the aorta (TRA) at the isthmus according to the following scheme:

1. Acute-less than 8 days after injury, with:
   1. Continuing free hemorrhage, or
      
   2. Periaortic hematoma contained within the mediastinum.
      
   
   
2. Chronic-pseudoaneurysm present more than 1 week after injury.
   

To avoid further confusion in the literature, the present report includes only acute (class I-A and I-B) ruptures at the isthmus.

	Material and Methods

Top Footnotes Abstract Introduction Material and Methods References

 
Subjects
We reviewed the medical and transportation records of all patients with an established diagnosis of traumatic rupture of the aorta at its isthmus resulting from injury occurring within 7 days of admission and encountered between 1964 and January 1, 1993. Patients who died during hospital resuscitation, the performance of diagnostic studies, or interhospital transportation were included. Patients whose tear involved the aortic arch proximal to the subclavian artery (n = 3) or near the diaphragm (n = 2), who arrived dead, and 2 patients operated on for a mistaken diagnosis were excluded. One hundred ten patients, 99 of whom underwent surgical repair of the aorta, met these criteria: 59 before July 1, 1984 [1], and 51 patients since then. Most patients (82.7%) were male and ranged in age from 15 to 87 years (mean, 33 years). Most patients (102/108, or 94.4%) had been involved in a motor vehicle accident. The time that elapsed from the injury to admission ranged from 0.5 to 48.0 hours (mean, 3.3 hours; median 1.5 hours). A general policy, consisting of early diagnosis made on the basis of aortographic findings and prompt repair performed with the patient under partial cardiopulmonary bypass, prevailed throughout.

Management Policies
An early chest radiogram (obtained in the anteroposterior view on a portable machine) is obtained in the receiving area in all victims of either blunt chest trauma or decelerating trauma (without evidence of chest injury). Aortography is done when any mediastinal abnormality is seen on the radiogram. Therapy is given priority when an injury appears imminently life-threatening, such as in the event of airway obstruction, external or intraabdominal hemorrhage, or continuing intracranial bleeding. In general, when the need for laparotomy is urgent, aortography is carried out after laparotomy. When the mediastinal radiographic findings are abnormal, a presumptive diagnosis of major vascular injury is made and hemodynamic monitoring and medical therapy are instituted before aortography is done; this was necessary in 44 of the last 51 patients.

Medical therapy includes the intravenous infusion of a vasodilator (usually nitroprusside), attempted limitation of the intravenous administration of fluids once the blood pressure exceeds 90 mm Hg, and the administration of a ß-blocker when the pulse rate exceeds 85 to 90 beats/min. Medical management is continued while diagnostic studies or other surgical procedures are being performed, or until the patient with proven rupture has been placed on cardiopulmonary bypass.

In 10 patients, aortic repair was electively delayed for days to weeks because of the presence of other lesions that were thought to make anesthesia and thoracotomy more risky than continued medical therapy.

Operation
Partial cardiopulmonary bypass was used in all patients since 1978, except for 2 in whom the clamp-and-sew technique was used to control massive intrathoracic hemorrhage before a pump-oxygenator could be available (about 15 to 30 minutes). Resident surgeons did all or a major part of the anastomoses in most cases. Single-lung ventilation was used when feasible. Cervical spine fractures, pulmonary injuries, obesity, fluid overload, and other factors contributed to causing problems in this area.

Femoral arterial and iliac venous exposure and a posterolateral fourth interspace or rib resection thoracotomy are started simultaneously. Three 5-cm-long segments of woven Dacron vascular prostheses (18, 20, and 22 mm) are kept packaged together in the operating room for use in this procedure and are preclotted before the administration of heparin. Control of the arch by tapes is made easier by opening the pericardium and dissecting the concavity from within. The external iliac vein is cannulated with a large (28F) catheter passed so that its tip is in or near the right atrium. After obtaining proximal and distal control of the aorta, when massive hemorrhage is encountered or when the surgical dissection becomes risky, PCPB is begun. Pump flow rates are adjusted to achieve a right radial pressure of 80 to 130 mm Hg, systolic and a mean distal perfusion pressure of more than 60 mm Hg, at a flow rate of more than 1,200 mL/min (usually 1,800 to 2,800 mL/min). The arterial pressures are kept at desirable levels by adding or removing blood from the system. All vasodilators are stopped with aortic clamping, as the spinal cord perfusion pressure may decrease as the resistance of the body below the clamp declines and the collateral circulation shunts blood away from the cord [2]. Attempts are made to avoid ligating intercostal arteries; back-bleeding may be prevented by temporarily occlusing any arteries within the clamped segment. The transection is completed, the aorta debrided, and an arterial prosthesis interposed with continuous 4-0 polypropylene sutures. Pump flow is continued until the proximal clamp is fully removed. The rate of removal depends upon the radial arterial pressure, which is maintained above 90 mm Hg, systolic. The pump prevents declamping shock and acts as a ``safety net'' to prevent the proximal aortic pressure from decreasing below the pump perfusion pressure during clamp removal.

Data Handling
Because not all relevant information is available in all records, different denominators are used in different data sets. Survival is reported at discharge and for 30 days after aortic repair (or injury, when repair was not done). As only 5 of the 110 patients operated on became paraplegic after repair and the characteristics of the groups of patients undergoing different operations are quite different, a meaningful statistical analysis of the data is unrewarding.

Results
Preoperative Status
Of the 110 patients, 95 (86.4%) had serious associated injuries. Half of these had sustained long bone fractures (50/95, 52.6%) or intraabdominal injury (49/95, 51.6%), or both. Central nervous system or serious facial injuries, or both, occurred in 71 (74.7%), and bilateral pulmonary contusions occurred in 27 (28.4%). Pelvic fractures (20.0%), cardiac contusions (9.5%), traumatic amputations (6.3%), and other lesions were less common. About a third of the patients (32.7%) had a normal arterial blood pressure on arrival; almost half (49.0%) were in a hemodynamically unstable condition, with a blood pressure of 0 to 70 mm Hg. Hypertension (>140 mm Hg) was present in 18.4%.

Aortography was done from 2.7 to 71.0 hours after injury (mean, 14.1 hours; median, 8.3 hours). During this time, many underwent laparotomy, craniotomy, or orthopedic procedures. The time that elapsed from the injury to aortic repair ranged from 4.2 to 68.3 hours (median, 9.1 hours), excluding 10 patients in whom late repair was elected. Free rupture of the periaortic hematoma occurred during the delay in 1 of these patients.

There were three treatment groups: those who died before diagnosis could be made or definitive treatment instituted (n = 8), those who received only medical therapy (n = 3), and those who had aortic repair (n = 99).

Operations
Although all patients treated surgically are comparable from the standpoint of having undergone some type of repair (Table 1), the groups are not comparable in other regards. Some of the patients who underwent clamp-and-sew procedures were operated on as a desperate attempt to control hemorrhage, frequently during efforts at resuscitation; the patients who were treated medically until their status improved had other injuries that were thought likely to be fatal in the event of further general anesthesia and surgery. Of the 44 recent patients who had aortic repair, 35 (79.5%) had other major surgical procedures (n = 51) immediately before (18 patients) or after (17 patients) aortic repair. In general, laparotomies were done before aortic repair and orthopedic procedures were done afterward. Intercostal arteries were ligated only when avulsed or involved in the suture line. In most cases (94.1%), none or only one was ligated.

Late operations included drainage of a hemothorax (or decortication), performed in 8 patients (19.5%); tracheotomy, performed in 6 (14.6%); and debridement and closure of a surgical thoracotomy, performed in 2, and of a groin wound infection, performed in 2 patients. A single patient each had an inferior vena caval filter installed, gastric resection performed, or an arm amputated because of phlegmasia.

Effects of Heparin
In 29 (33.0%) of the 88 patients who had repair performed under PCPB, a serious head injury was suspected before the thoracotomy was carried out; however, 10 of these were later found to be normal. The intracranial injury worsened after aortic repair in 3 patients (15.8%). The conditions of 3 other patients with open cerebral wounds or skull fractures associated with basal ganglia injury did not deteriorate after aortic repair. Another patient with hemiplegia, basilar skull fracture, multiple brain infarcts, and a fracture at C4 died of adult respiratory distress syndrome 15 days after aortic repair, after an improvement in his neurologic status. Thoracotomy had been delayed in this patient because of the need to perform a laparotomy, and because of the presence of coagulopathy, a fever that exceeded 39.4°C, and serious head and cervical spinal injuries. In none of the 4 patients with spinal injuries who underwent aortic repair with PCPB did new neurologic abnormalities develop postoperatively.

Pulmonary injury, associated with severe contusion, flail chest, and penetration with rib fragments, appeared to worsen during the operation in 6 patients. The problems included increased bleeding, expansion of areas of hemorrhagic discoloration, air leakage, and iatrogenic trauma incurred during retraction. Two had to undergo lobectomy for control of bleeding and air leaks during operation. In 5, single-lung ventilation was unsatisfactory due to technical factors.

Evidence of increased or new intraabdominal bleeding arising during PCPB was not observed, and none died as the result of intraabdominal bleeding after aortic repair. Extensive soft tissue injuries bled during the period of heparinization, but this loss was offset by the marked reduction in blood loss from the chest brought about by the recirculation of shed blood.

Paraplegia
Among the 99 patients who had aortic repair, 5 (5.1%) were paraplegic after operation (see Table 1). Paraplegia developed in 2 of 86 evaluable patients (2.3%) who underwent repair with PCPB; both had exhibited severe hypoxia of the proximal body during bypass. One patient with fractures at T10 and T11 was paraplegic before operation. The total aortic cross-clamp times ranged from 18 to 82 minutes (mean, 47.2 minutes; median, 45 minutes). The 3 patients who suffered paraplegia after undergoing the clamp-and-sew technique had cross-clamp times of 40, 43, and 45 minutes; of the 2 who suffered paraplegia after PCPB, the cross-clamp times were 52 and 81 minutes. Four of the 5 patients who suffered paraplegia were among those in the early series of 59 patients, and this was associated with the clamp-and-sew technique in 3 and with PCPB in 1. All patients experienced periods of hypotension or severe prolonged hypoxemia during the preoperative or operative periods, or at both times.

Only 1 patient in the recent group suffered paraplegia, an 18-year-old male patient who had sustained subluxation of C2 and C3. He was operated on for a fractured patella, after which findings from a chest computed tomographic scan prompted the performance of aortography. The aorta was repaired 14 hours after admission. The cross-clamp time was 81 minutes, and the time was prolonged due to problems in ventilation: the endotracheal tube became displaced and reinsertion was complicated by his cervical spine injury. He was in a hemodynamically stable condition, with a short period of hypoxia (oxygen tension, 69 mm Hg) before repair. His oxygen tension ranged from 68 to 70 mm Hg during much of the operation. He received 0.25 mg and 0.5 mg of propanolol for the control of hypertension during the early part of the procedure, followed by a continuous drip of nitroglycerin throughout the aortic cross-clamp period. The pump flow to the distal body was about 2,000 mL/min. He awoke with T6-7 paraplegia. This led to our adopting the policy of not administering vasodilators during the cross-clamp period and of controlling the proximal blood pressure by the pump circuit alone. There have been no instances of paraplegia in the subsequent 39 patients.

Comment
A retrospective and uncontrolled clinical series such as this has scientific limitations. We have restricted the case selection to a clearly described type of pathologic condition and, since 1984, have adhered to a predetermined policy and technique of management and operation, as well as to a specific method of prospectively observing neurologic and other complications.

About 8% to 16% of the victims of motor vehicle accidents who undergo autopsy are found to have ATRA [3]. We confirmed rupture of the aortic isthmus in 51 patients among 24,681 consecutively admitted patients with major multisystem trauma, with 12,587 of the cases resulting from motor vehicle accidents (a rate of about four ruptures per 1,000 cases of trauma). Injuries of the ascending or transverse aorta are usually associated with severe thoracic skeletal and cardiac injury, whereas ruptures at the isthmus frequently occur in the absence of other major intrathoracic injuries. We believe that the mechanism of injury (crush versus deceleration) is different and that this accounts for the high incidence of death in victims with nonisthmus rupture before they reach the hospital. At least half of the patients with ATRA die before reaching a surgical facility, usually from associated injuries, such as head, cervical spine, cardiac, or intraabdominal hemorrhage [3–7].

Diagnosis and Initial Management
Any abnormality of the mediastinum on the chest radiogram should prompt a presumptive diagnosis of great vessel injury. Experience has convinced us of the fallacy of depending on computed tomography to clearly define the nature of an aortic injury, an observation also made by others [8–10]. Newer techniques of computed tomographic scanning (eg, helical and spiral) appear to yield much better images of the aorta and may prove to be as accurate in defining the nature of the injury as aortography. We have had three false-positives aortographic results indicating TRA at the isthmus, but no known false-negative findings, among more than 400 aortograms obtained when great vessel injuries were suspected. In the past 2 years, we have employed transesophageal echocardiography to delineate the aorta before performing aortography, for example, during laparotomy for the control of hemorrhage. Although very promising [11], the technique has not yet been adequately evaluated in a large series of patients to ascertain its merits. Our policy of instituting medical therapy after an abnormal mediastinum is seen on x-ray films means that, when the pseudoaneurysm is contained (class I-B), pharmacologic intervention to mitigate aortic wall stresses is started immediately, allowing aortic repair to be delayed until more urgent problems are handled. Two of the 44 patients treated using this protocol suffered fatal rupture of the hematoma after arrival at the hospital, in 1 during (prompt) aortography before the diagnosis was established and in 1 during laparotomy for the control of massive hemorrhage. In both patients, the arterial blood pressure was more than 180 mm Hg near the time of rupture. Other investigators have also noted the lack of correlation between a delay in diagnosis or operation and patient outcome [9, 12, 13].

In 1991, several European surgeons reported poorly defined groupings of acute and chronic lesions [7, 12, 14]. Our classification seems to be more clinically useful and to allow more uniformity in reporting.

Anesthesia
Arterial hypotension or hypoxemia, or both, developing before, immediately after, or especially during aortic cross-clamping is associated with an increased probability of paraplegia. Failure of single-lung ventilation is influenced by intubation of the patient en route to the hospital or by the presence of a concomitant spinal injury. Single-lung ventilation greatly improves exposure of the aorta, prevents blood from an injured left lung from draining into the right lung during lateral thoracotomy, and ameliorates the iatrogenic pulmonary injury incurred during surgical retraction.

Nitroprusside or nitroglycerin, sometimes given during anesthesia to decrease the proximal hypertension of aortic cross-clamping, is considered contraindicated because a peripheral vasodilating effect below the aortic cross-clamp could cause blood ``steal'' from the cord. Although nitroprusside does lower the elevated central venous pressure, which interferes with cord blood flow, it also lowers arterial pressure to the cord and does not lower the elevated cerebrospinal fluid pressure, which also causes the cord blood flow to be reduced.

Paraplegia
The incidence of paraplegia after the repair of a TRA in human subjects is reported to range from 3% to 33%. This variation in incidence is primarily due to the different clinical characteristics of the patient populations reported on from various hospitals. For example, only 4 of the 51 patients (7.8%) in Hilgenberg and colleagues' [9] series arrived directly from the accident scene without being transferred from another hospital, while over 60.2% of Cowley and associates' patients [4] and 51.0% of ours arrived directly from the accident scene.

The overall incidence of paraplegia, including many unpublished cases, is probably around 25% for the patients undergoing surgical repair [15]. Cord damage that occurs after aortic cross-clamping during repair of TRA is usually the result of ischemia during the cross-clamp period. Svensson and Crawford's [16] excellent work delineating the pathophysiologic characteristics of ischemic myelopathy has pointed up how the reperfusion period as well as the variety of factors that may be involved produce this ischemia. Although such etiologic factors as an anomalous origin of the anterior spinal artery, traumatic disruption of large numbers of intercostal arteries, and other hypotheses have been suggested, there is no published objective evidence that the paraplegia occurring in patients operated on for the repair of a TRA is related to these circumstances. A case of paraplegia resulting from repair of a TRA in which an anomalous anterior spinal artery has been found at autopsy or demonstrated by angiography has not yet been reported. Such factors are responsible for causing other conditions, such as the anterior spinal artery syndrome that develops after damage incurred during a spinal surgical procedure or during permanent division of multiple intercostal arteries in aneurysm resections. The findings from extensive experimental studies performed in rats, rabbits, cats, sheep, dogs, pigs, and baboons, and involving the use of various protocols, have confirmed the concept that prolonged (±30 minutes) ischemia alone is sufficient to consistently produce ischemic myelopathy.

With simple clamping of the aorta, the duration of aortic cross-clamping is clearly related to the incidence and degree of ischemic myelopathy [4, 9, 10, 17]. Zieger and associates [17] reported that no paraplegia occurred among their 11 patients whose cross-clamp times had been less than 28 minutes, but paraplegia arose in all 7 patients whose times had exceeded 35 minutes, a finding that was significant at p < 0.002. In larger series from both Boston [9] and Maryland [4], no myelopathy developed in those patients whose cross-clamp times had been less than 30 minutes, but there was a disturbing incidence when the times were longer. The patients who suffered myelopathy in Wallenhaupt and colleagues' [10] series had a mean cross-clamp time of 48 minutes, compared with 26 minutes in those who did not suffer cord damage (p 0.05). Other factors, such as shock and hypoxia, contribute to the probability of cord damage. Cross-clamp periods of 30 minutes or less in hemodynamically stable patients without hypoxia are rarely associated with cord damage.

Cord Protection
Following are some of the different techniques used to protect the cord during aortic cross-clamping:

• Hypothermia: systemic, local perfusion of the cord.
  
• Shunts
  • Passive: distal perfusion depends on the left ventricular pressure. The shunt extends from the ventricle or ascending aorta to the femoral artery or descending aorta.
    
  • Active: mechanical pump interposed between the inlet and outlet. The shunt extends from the atrium, ventricle, or ascending aorta to the femoral artery or descending aorta.
    
  
  
• Cardiopulmonary bypass: with oxygenator.
  • Partial: from the right ventricle (via the iliac vein or veins) to the aorta or femoral artery.
    
  • Total.
    
  
  
• Physiologic: cerebrospinal fluid drainage, hypothermia, and so on.
  • Pharmacologic treatments
    
  • Intrathical administration of papaverine.
    
  • Calcium channel blockers.
    
  • Allopurinal.
    
  • Superoxide dismutase.
    
  • Perfusion of cord with cold ``neuroplegic'' solution.
    
  
  

The clamp-and-sew method does not give any protection to the spinal cord. Many of the innovations just listed have demonstrated, at least in animals, an ability to offer some protection to the spinal cord during aortic cross-clamping. Hypothermia and various arterial bypass grafts were used for aneurysm resections in the 1950s. Left atrial-to-femoral artery pumping was found to be unsatisfactory, primarily because of the systemic heparinization needed. Heparin-bonded (Gott) shunts that extended from the proximal to the distal aorta were introduced to overcome these dangers. However, early experiences with each of these methods were generally unsatisfactory; the aorta was frequently repaired in patients in unstable conditions, many with active intraabdominal bleeding. The good experiences with aneurysm resection involving the use of simple cross-clamping led to the adoption of the clamp-and-sew technique for the repair of an ATRA. However, the incidence of paraplegia after this procedure remains unsatisfactory. The patient with ATRA poses far more pathophysiologic challenges than does the patient with an aneurysm presenting for elective repair; these problems include pulmonary and cardiac contusions, shock, hypoxia, hypoosmolality resulting from fluid overload, and abrupt disruption of cord blood flow in the absence of preformed collaterals.

There has never been a valid comparison of the outcome from several surgical techniques in matched groups of patients treated in the same time period. Table 3 summarizes the findings from relatively large recent series in which adequate information is available for estimating the risks of repair of a TRA, and, on the basis of the findings reported, the danger of the clamp-and-sew technique appears obvious. Others have also pointed out that ``the clamp/sew technique was inadequate when multiple aortic tears were found intraoperatively'' [13]. Unfortunately, the details furnished in most published reports are not specific enough to allow clear conclusions to be drawn. Patients whose repair consisted of various types of passive shunts from different proximal sites to various distal sites are frequently combined into the same data set, and these shunts are called either ``shunts'' or ``bypasses''; active shunts using an extracorporeal pump, without an oxygenator, (``left heart bypass'') from the same various sites also may be included. Some series also include in such groups patients who have had true cardiopulmonary bypass involving the use of an oxygenator. Obviously, these techniques produce very great physiologic differences, and such lumping of patients does not lead to valid conclusions. The type, size, and technique of the shunt used are frequently inadequately described, and there is rarely evidence that the shunt maintained adequate distal flow or pressure. Longer cross-clamp times in the shunted groups may reflect more extensive injuries, and shunts have commonly been used after the clamp-and-sew technique proved unsatisfactory and was abandoned. As pointed out by the Massachusetts General Hospital group [9], ``the length of the cross-clamp time is mainly a reflection of the complexity of the injury and the degree of difficulty of the repair. . . .'' In most series, cross-clamp times are longer in patients undergoing bypass repair than in those undergoing clamp-and-sew repair.

Active shunts, such as the vortex blood pump, with heparin-bonded circuits placed from the left heart or proximal aorta, allow for controlled perfusion of the distal body without the need for systemic heparinization. The results from the use of this technique in patients with ATRA have been encouraging. No cord damage was noted among a total of 70 patients in whom this technique was used [8, 17, 19–21]. However, this technique does not allow for support of the circulation in the event of a failing heart, adequate oxygenation in the presence of pulmonary damage, or the reinfusion of shed blood.

Although the disadvantages of PCPB and heparin have been repeatedly mentioned in the literature, we can find no objective or recent (15 years) evidence that the disadvantages outweigh the advantages. The high mortality rate associated with aortic repair observed in the past decades can be misleading. For example, Merrill and associates [3] reported a mortality rate of 28.6% with PCPB, but, of their total of 7 patients who underwent repair, 1 died of a ruptured right ventricle and the other patient, who had severe shock and acidosis before operation, died of pulmonary edema.

Both of our patients who became paraplegic after repair that involved the use of PCPB had profound hypoxia of the proximal body caused by pulmonary problems during the cross-clamp period. Other observers also report a low incidence of paraplegia associated with the use of PCPB (see Table 3). Statistical analysis or meta-analysis of these data is not feasible because there is too much variation in the patients' characteristics and the reported data. The table represents a reasonable summary of the material reported. Partial cardiopulmonary bypass has been shown to yield very low rates of paraplegia [3, 8, 11, 22]: of 31 patients surviving aortic repair under PCPB reported on by Kodali and colleagues [6], 1 became paraplegic (3.2%), versus 4 who suffered paraplegia among the 8 patients who survived repair with the clamp-and-sew technique. Following are some of the advantages and disadvantages of PCPB in the repair of ATR at the aortic isthmus:

• Advantages
  • Decreases incidence of cord damage by decreasing elevated central venous pressure and elevated cerebrospinal fluid pressure
    
  • Protects the left ventricle and lung vascular bed
    
  • Expedites reinfusion of shed blood
    
  • Decreases time constraints of aortic cross-clamping
    
  • Decreases ``declamping shock''
    
  • Aids in control of hypothermia
    
  
  
• Disadvantages
  • Requires heparinization in presence of trauma
    
  • Requires pump oxygenator, perfusionist, cannulation of groin vessels
    
  • Awkward positioning when there is major pelvic or lower extremity injury
    
  • Appears to increase existing lung damage
    
  • Contraindicated in the presence of a serious intracranial injury
    
  • (may increase bleeding or coagulopathy?)
    
  
  

Heparin Treatment
During the 1960s and 1970s, most patients with TRA underwent thoracotomy as quickly as possible after limited diagnostic studies were performed, before they underwent operation for the control of intraabdominal or intracranial bleeding. The mortality in these patients was high, with ``hemorrhage'' cited as the primary cause of death, and heparin therapy was usually considered contraindicated. There are little data to support this contention, however. Better prioritization with attention to the control of critical bleeding, especially within the abdomen, before aortic repair has led to the prevention of fatal hemorrhage in patients with class I-B ruptures. Bouchart and co-workers [22], in a report on 41 patients undergoing repair of TRA with PCPB, observed that ``heparinization did not increase the risk when orthopedic or abdominal lesions were treated before the aortic lesion.'' Guvendik and associates [8], based on their observations in 21 cases, also believe that the dangers of systemic heparinization are exaggerated. Systemic heparin treatment was used by Fabian and colleagues [23] in a large group of seriously injured trauma patients who had sustained blunt carotid injuries, without any major bleeding complications arising.

The neurologic status of 3 of our patients deteriorated during operation, but this incidence does not appear to be higher than that seen in other patients with comparable head injuries having laparotomy or thoracotomy without heparin treatment. We have been unable to demonstrate a deleterious effect of heparin treatment on the neurologic outcome; others agree [9, 11, 14, 17, 20, 22]. We consider the presence of brain tissue extruding from the skull or evidence of active intracranial bleeding or a space-occupying intracranial hematoma to be contraindications to immediate aortic repair and indications for the institution of medical therapy for the management of a TRA while the patient's neurologic status is stabilized.

An increase in pulmonary abnormalities was obvious in a few of our patients with traumatic pulmonary contusion and lacerations preoperatively, who suffered iatrogenic trauma during retraction of the lung during heparinization. This phenomenon was not observed in association with satisfactory single-lung anesthesia.

Considering the experience recorded in the literature and our own experience, there is no recent or objective evidence that the disadvantages of heparin treatment and PCPB outweigh the important advantages.

Other Techniques and Future Trends
Techniques for preventing paraplegia include various physiologic and pharmacologic manipulations. Thirty years ago, Blaisdell showed that the drainage of cerebrospinal fluid during aortic cross-clamping protected neurologic function in animals; this idea has recently been resurrected [24]. Cooley has more recently used the ``open distal aorta'' method for decreasing cerebrospinal fluid pressure during aneurysm repairs.

Svensson and Crawford [16] demonstrated that the intrathecal administration of papaverine before aortic cross-clamping is quite effective in preventing myelopathy in baboons, and this was also observed in 11 operations on human subjects. Calcium channel blockers, allopurinol, superoxide dismutase, barbiturates, and perfusion of the cord with cold crystalloid solution and with dismutase and thiopental have also proved beneficial [16, 25]. Paraplegia may result from the anterior spinal artery syndrome or the occlusion of multiple intercostal arteries stemming from trauma to the spine, especially when there is preexisting spinal disease or the traumatic involvement of multiple intercostal arteries. These rare instances of ischemic myelopathy stemming from another injury associated with TRA result in an irreducible incidence of paraplegia in severely traumatized patients.

Future improvements in the management of TRA will include the wider use of transesophageal echocardiography and better techniques of computed tomography in its diagnosis, the institution of medical therapy before transfer of the patient to a trauma center (and before aortography?), elective delay in the repair of more patients with clarification of the indications for immediate repair, and newer techniques (such as pumping without heparin, intraluminal shunts, local hypothermia, and drugs) for preventing paraplegia.

In conclusion, we have demonstrated that the spinal cord protection conferred by adequate perfusion of the lower body with oxygenated blood by PCPB results in a low incidence of paraplegia. Heparin therapy is not contraindicated if other severe hemorrhage is controlled before aortic repair and an injured left lung is protected by being collapsed during the repair. The optimal approach to the prevention of ischemic myelopathy will continue to evolve, and probably will consist of a combination of pharmacologic (physiologic) and bypass techniques.

	Footnotes

Top Footnotes Abstract Introduction Material and Methods References

 
Address reprint requests to Dr Pate, Department of Surgery, University of Tennessee-Memphis, 956 Court Ave, Memphis, TN 38163.

	References

Top Footnotes Abstract Introduction Material and Methods References

 

1. Pate JW. Traumatic rupture of the aorta: emergency operation. Ann Thorac Surg 1985;39:531–7.[Abstract]
2. Laschinger JC, Owen J, Rosenbloom M, Cox JL, Kouchoukos NT. Detrimental effects of sodium nitroprusside on spinal cord motor tract perfusion during thoracic aortic cross-clamping. Surg Forum 1987;38:195–6.
3. Merrill WH, Lee RB, Hammon JW Jr, Frist WH, Stewart JR, Bender HW Jr. Surgical treatment of acute traumatic tear of the thoracic aorta. Ann Surg 1988;207:699–706.[Medline]
4. Cowley RA, Turney SZ, Hankins JR, Rodriguez A, Attai S, Shankar BS. Rupture of thoracic aorta caused by blunt trauma. A fifteen year experience. J Thorac Cardiovasc Surg 1990;100:652–61.[Abstract]
5. Duhaylongsod FG, Glower DD, Wolfe WG. Acute traumatic aortic aneurysm: the Duke experience from 1970 to 1990. J Vasc Surg 1992;15:331–43.[Medline]
6. Kodali S, Jamieson WR, Leia-Stephens M, Miyagishima RT, Janusz MT, Tyers GF. Traumatic rupture of the thoracic aorta. A 20-year review: 1969-1989. Circulation 1991;84 (Suppl 5):11140–6.
7. Regensburger D, Sievers HH, Kraatz EG, Uhlmann J. Traumatische Aortenruptur: Sofortoperation-Intervalloperation? Langenbecks Arch Chir [Suppl Kongressbd] 1991:558-62.
8. Guvendik L, Davis NR, Starr A. Repair of traumatic aortic transection: a management protocol and review of twenty-one patients. Thorac Cardiovasc Surg 1988;36:198–201.[Medline]
9. Hilgenberg AD, Logan DL, Akins CW, et al. Blunt injuries of the thoracic aorta. Ann Thorac Surg 1992;53:233–9.[Abstract]
10. Wallenhaupt SL, Hudspeth AS, Mills SA, Tucker WY, Dobbins JE, Cordell AR. Current treatment of traumatic aortic disruptions. Am Surg 1989;55:316–20.[Medline]
11. Kieny R, Carpentier A. Traumatic lesions of the thoracic aorta: a report of 73 cases. J Cardiovasc Surg 1991;32:613–9.[Medline]
12. Stulz P, Reymond A, Bertschmann W, Graedel E. Decision-making aspects in the timing of surgical intervention in aortic rupture. Eur J Cardiothorac Surg 1991;5:623–7.[Abstract]
13. Cernaianu AC, Cilley JH Jr, Baldino WA, Spence RK, DelRossi AJ. Determinants of outcome in lesions of the thoracic aorta in patients with multiorgan system trauma. Chest 1992;101:331–5.[Abstract/Free Full Text]
14. Fraedrich G, Spillner G, Schlosser V. Die frische und veraltete Aortenruptur. Versorgung ohne und mit dem Schutz eines Bypassverfahrens. Langenbecks Arch Chir [Suppl Kongressbd] 1991:571-5.
15. Pate JW. Traumatic rupture of the aorta. A survey of 40 North American surgeons' experiences. In: Proceedings of the 1988 World Congress of Surgery, Sydney, Australia.
16. Svensson LG, Crawford ES. Aortic dissection and aortic aneurysm surgery: clinical observations, experimental investigations, and statistical analyses. Part I. Curr Probl Surg 1992;29:817–911.[Medline]
17. Zieger MA, Clark DE, Morton JR. Reappraisal of surgical treatment of traumatic transection of the thoracic aorta. J Cardiovasc Surg 1990;31:607–10.[Medline]
18. Verdant A. Descending thoracic aorta aneurysms: surgical treatment with the Gott shunt. Can J Surg 1992;35:493–6.[Medline]
19. Benckart DH, Magovern GJ, Liebler GA, et al. Traumatic aortic transection: repair using left atrial to femoral bypass. J Cardiac Surg 1989;4:43–9.[Medline]
20. Higgins RS, Sanchez JA, DeGuidis L, et al. Mechanical circulatory support decreases neurologic complications in the treatment of traumatic injuries of the thoracic aorta. Arch Surg 1992;127:516–9.[Abstract]
21. Read RA, Moore EE, Moore FA, Haenel JB. Partial left heart bypass for thoracic aorta repair: survival without paraplegia. Arch Surg 1993;128:746–52.[Abstract]
22. Bouchart F, Bessou JP, Tabley A, et al. Traumatic isthmic ruptures of the aorta during the acute phase. Ann Chir 1992;46:116–24.[Medline]
23. Fabian T, George SM Jr, Croce MA, Mangiante EC, Voeller GR, Kudsk KA. Carotid artery trauma: management based on mechanism of injury. J Trauma 1990;30:953–63.[Medline]
24. McCullough JL. Paraplegia after thoracic aortic occlusion: influence of cerebrospinal fluid drainage. J Vasc Surg 1988;7:153–60.[Medline]
25. Kirschner DL, Kirshner RC, Heggeness LM, DeWeese JA. Spinal cord ischemia: an evaluation of pharmacologic agents in minimizing paraplegia after aortic occlusion. J Vasc Surg 1989;9:305–8.[Medline]
26. Schmidt CA, Wood MN, Razzouk AJ, Killeen JD, Gan KA. Primary repair of traumatic aortic rupture: a preferred approach. J Trauma 1992;32:588–92.[Medline]

This article has been cited by other articles:

				D. Veerasingam, M. Vioreanu, M. O' Donohue, and J. F. McCarthy Traumatic transection of the innominate artery Interactive CardioVascular and Thoracic Surgery, December 1, 2003; 2(4): 569 - 571. [Abstract] [Full Text] [PDF]

				G. M. Hochheiser, D. E. Clark, and J. R. Morton Operative Technique, Paraplegia, and Mortality After Blunt Traumatic Aortic Injury Arch Surg, April 1, 2002; 137(4): 434 - 438. [Abstract] [Full Text] [PDF]

				T. Langanay, J.-P. Verhoye, H. Corbineau, A. Agnino, T. Derieux, P. Menestret, Y. Logeais, and A. Leguerrier Surgical treatment of acute traumatic rupture of the thoracic aorta a timing reappraisal? Eur. J. Cardiothorac. Surg., February 1, 2002; 21(2): 282 - 287. [Abstract] [Full Text] [PDF]

				E. A. Hessel Bypass Techniques for Descending Thoracic Aortic Surgery Seminars in Cardiothoracic and Vascular Anesthesia, November 1, 2001; 5(4): 293 - 320. [Abstract] [PDF]

				S. W. Downing, M. G. Cardarelli, J. Sperling, S. Attar, D. C. Wallace, A. Rodriguez, J. Brown, G. J. R. Whitman, and J. S. McLaughlin Heparinless partial cardiopulmonary bypass for the repair of aortic trauma J. Thorac. Cardiovasc. Surg., December 1, 2000; 120(6): 1104 - 1111. [Abstract] [Full Text] [PDF]

				A. J. Razzouk, S. R. Gundry, N. Wang, M. J. del Rio, D. Varnell, and L. L. Bailey Repair of Traumatic Aortic Rupture: A 25-Year Experience Arch Surg, August 1, 2000; 135(8): 913 - 918. [Abstract] [Full Text] [PDF]

				E. Tatou, E. Steinmetz, S. Jazayeri, B. Benhamiche, R. Brenot, and M. David Surgical outcome of traumatic rupture of the thoracic aorta Ann. Thorac. Surg., January 1, 2000; 69(1): 70 - 73. [Abstract] [Full Text] [PDF]

				A. T. Gurbuz, D. S. Weiman, and J. W. Pate Traumatic injury to the thoracic aorta Ann. Thorac. Surg., September 1, 1999; 68(3): 1116 - 1117. [Full Text] [PDF]

				J. S. Gammie, A. S. Shah, B. G. Hattler, R. L. Kormos, A. B. Peitzman, B. P. Griffith, and S. M. Pham Traumatic aortic rupture: diagnosis and management Ann. Thorac. Surg., October 1, 1998; 66(4): 1295 - 1300. [Abstract] [Full Text] [PDF]

				J. W. Pate Modern management of traumatic rupture of the aortic isthmus Ann. Thorac. Surg., August 1, 1998; 66(2): 611 - 612. [Full Text] [PDF]

				M. S. Sweeney, D. J. Young, O. H. Frazier, P. R. Adams, M. O. Kapusta, and M. P. Macris Traumatic Aortic Transections: Eight-Year Experience With the "Clamp-Sew" Technique Ann. Thorac. Surg., August 1, 1997; 64(2): 384 - 387. [Abstract] [Full Text]

				G. D. Trachiotis, J. E. Sell, G. D. Pearson, G. R. Martin, and F. M. Midgley Traumatic Thoracic Aortic Rupture in the Pediatric Patient Ann. Thorac. Surg., September 1, 1996; 62(3): 724 - 731. [Abstract] [Full Text]

				A. C. Nicolosi, G. H. Almassi, M. Bousamra II, G. B. Haasler, and G. N. Olinger Mortality and Neurologic Morbidity After Repair of Traumatic Aortic Disruption Ann. Thorac. Surg., March 1, 1996; 61(3): 875 - 878. [Abstract] [Full Text]

This Article Abstract Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Pate, J. W. Articles by Walker, W. A. Search for Related Content PubMed PubMed Citation Articles by Pate, J. W. Articles by Walker, W. A.