Title: Spinal Cord Protection During Surgical Procedures on the Descending Thoracic and Thoracoabdominal Aorta
Abstract: arteria radicularis magna adenosine triphosphate cardiopulmonary bypass cerebrospinal fluid motor evoked potential polyethylene glycol spinal cord ischemia somatosensory evoked potential superoxide dismutase During the past 3 decades significant advances have been made in the surgical treatment of diseases affecting the aorta. Despite these important advances paraplegia remains a devastating complication of procedures on the thoracic and the thoracoabdominal aorta.1Crawford ES Coselli JS Thoracoabdominal aneurysm surgery.Sem Thorac Cardiovasc Surg. 1991; 3: 300-322PubMed Google Scholar, 2Kouchoukos NT Spinal cord ischemic injury: is it preventable.Sem Thorac Cardiovasc Surg. 1991; 3: 323-328PubMed Google Scholar, 3Svensson LG Loop FD Prevention of spinal cord ischemia in aortic surgery.in: Bergan JJ Yao JST Arterial surgery: new diagnostic and operative techniques. Grune & Stratton, New York1988: 273-285Google Scholar, 4Hollier LH Moore WM Avoidance of renal and neurologic complications following thoracoabdominal aortic aneurysm repair.Acta Chir Scand. 1990; 555: 129-135Google Scholar, 5Brewer LA Fosberg RG Mulder GA et al.Spinal cord complications following surgery for coarctation of the aorta.J Thorac Cardiovasc Surg. 1972; 64: a368-79PubMed Google Scholar, 6Mauney MC Blackbourne LH Lagenberg SE et al.Prevention of spinal cord injury after repair of the thoracic or thoracoabdominal aorta.Ann Thorac Surg. 1995; 59: 245-252Abstract Full Text PDF PubMed Scopus (99) Google Scholar Paraparesis and paraplegia occur as a direct result of interruption of blood flow to the spinal cord during these procedures. A number of techniques have been advocated for the prevention of spinal cord ischemic injury. This article will present our current understanding of the extent of this problem, the mechanism of injury, and the methods that have been devised to reduce the frequency of paraplegia after procedures on the descending aorta. Paraparesis and paraplegia can result after surgical procedures on the thoracic and abdominal aorta. The incidence of spinal cord injury seems to correlate with the extent of the aorta which is resected and whether the area of resection encompasses the peridiaphragmatic aorta. Table 1 summarizes the incidence of spinal cord ischemic injury (SCI). The frequency of SCI ranges from 0.2% after elective repair of an abdominal aortic aneurysm to as high as 40% in the setting of acute dissection or rupture involving the descending thoracic or thoracoabdominal aorta.7Svensson LG Crawford ES Hess KR et al.Experience with 1509 patients undergoing thoracoabdominal aortic operations.J Vasc Surg. 1993; 17: 357-370Abstract Full Text Full Text PDF PubMed Scopus (1084) Google Scholar, 8Lerberg DB Hardesty RL Siewers RD et al.Coarctation of aorta in infants and children: 25-year experience.Ann Thorac Surg. 1982; 33: 159-170Abstract Full Text PDF PubMed Scopus (99) Google Scholar, 9Duhaylongsod FG Glower DD Wolfe WG Acute traumatic aortic aneurysm: the Duke experience from 1970 to 1990.J Vasc Surg. 1992; 15: 331-343Abstract Full Text Full Text PDF PubMed Scopus (85) Google Scholar, 10Hess PJ Howe HR Robicsek F et al.Traumatic tears of the thoracic aorta: improved results using the Bio-Medicus pump.Ann Thorac Surg. 1989; 48: 6-9Abstract Full Text PDF PubMed Scopus (46) Google Scholar, 11Hilgenberg AD Logan DL Akins CW et al.Blunt injuries of the thoracic aorta.Ann Thorac Surg. 1992; 53: 233-239Abstract Full Text PDF PubMed Scopus (87) Google Scholar Similarly, in thoracoabdominal aortic aneurysms, SCI correlates with the length of the thoracic aorta which is involved.1Crawford ES Coselli JS Thoracoabdominal aneurysm surgery.Sem Thorac Cardiovasc Surg. 1991; 3: 300-322PubMed Google Scholar In Crawford type 4 thoracoabdominal aneurysms, where most or all of the abdominal aorta is replaced, SCI is seen in 1% of the patients. On the other hand, in Crawford Type 2 thoracoabdominal aneurysms, where most or all of the descending thoracic and abdominal aorta is replaced, the incidence of SCI is as high of as 21%. The risk of SCI is increased significantly in the presence of a dissection in the setting of a thoracoabdominal aneurysm, perioperative hypotension, long period of aortic cross-clamping, increased cerebrospinal fluid (CSF) pressure, sacrifice of critical intercostal or lumbar arteries, and the extent of aortic disease.12Crawford ES Crawford JL Safi HJ et al.Thoracoabdominal aortic aneurysms: preoperative and intraoperative factors determining intermediate and long-term results in 605 patients.J Vasc Surg. 1986; 3: 389Crossref PubMed Google Scholar, 13Cox GS O'Hara PJ Hertzer NR et al.Thoracoabdominal aneurysm repair: a representative experience.J Vasc Surg. 1992; 15: 780-788Abstract Full Text Full Text PDF PubMed Scopus (196) Google Scholar, 14Livesay JL Cooley DA Ventemiglia RA et al.Surgical experience in descending thoracic aneurysmectomy with and without adjuncts to avoid ischemia.Ann Thorac Surg. 1995; 39: 37-46Abstract Full Text PDF Scopus (240) Google ScholarTable 1Frequency of Paraparesis or Paraplegia After Aortic Surger y, %Abdominal aortic aneurysm 0.2Ruptured abdominal aortic aneurysm2Aortic coarctation 0.4Chronic posttraumatic aneurysm<5Acute posttraumatic aneurysm 0-24Descending thoracic aortic aneurysm<10Thoracoabdominal aortic aneurysm10-20Acute dissection or rupture40 Open table in a new tab During embryologic development, 62 radicular arteries supply the spinal cord. Most of these arteries undergo degeneration in the embryo. By birth, 10 to 23 radicular arteries fuse to form the paired posterior spinal arteries. These vessels are continuous throughout the length of the spinal cord and supply blood to the posterior one third of the spinal cord. On the other hand, by birth a single anterior spinal artery is formed by 6 to 8 radicular arteries. This single vessel supplies the anterior two thirds of the spinal cord. There are no communicating branches between the anterior and posterior spinal arteries.15Sliwa JA Maclean IC Ischemic myelopathy: a review of spinal vasculature and related clinical syndromes.Arch Phys Med Rehabil. 1992; 73: 365-372Abstract Full Text PDF PubMed Scopus (94) Google Scholar, 16Svensson LG Klepp P Hinder RA Spinal cord anatomy of the baboon: comparison with man and implications of spinal cord blood flow during thoracic aortic cross-clamping.S Afr J Surg. 1986; 24: 32-34PubMed Google Scholar, 17Dommisse GF The blood supply of the spinal cord: a critical vascular zone in spinal surgery.J Bone Joint Surg (Br). 1974; 56: 225-235PubMed Google Scholar Unlike the continuous posterior spinal arteries, the anterior spinal artery is discontinuous and supplies the spinal cord in a segmental fashion. The segmental distribution of the anterior spinal artery is highly variable. The anterior spinal artery receives blood in a segmental fashion from the left vertebral artery, the thyrocervical trunk, the intercostal arteries, and the lateral sacral artery. Only a fraction of the segmental arteries reach the spinal cord. The superior cord to the level of T2 is supplied by three to five anterior radicular arteries, the mid-thoracic cord has up to one radicular artery and the lower thoracic, and lumbar spinal cord is supplied by three to five anterior radicular arteries including the arteria radicularis magna (ARM). The major radicular supply to the anterior spinal artery in the thoracic and upper abdominal region is provided by the ARM, known as the arteries of Adamkiewicz. Although in the majority of instances ARM originates from the descending aorta between the levels of T-9 and T-12, there is significant variability in the origin of this artery.18Djindjian R Hurth RM Houdart M et al.Arterial supply of the spinal cord.in: Djindjian R Hurth RM Houdart M Angiography of the spinal cord. University Park Press, Baltimore1970: 3-13Google Scholar, 19Hollier LH Causes and prevention of spinal cord ischemia.in: Veith FJ Current critical problems in vascular surgery vol 2 St. Quality Medical Publishing, Louis1990Google Scholar In 80% of instances ARM has a left-sided origin.20Svensson LG Rickards E Coull A et al.Relationship of spinal cord blood flow to vascular anatomy during thoracic aortic cross-clamping and shunting.J Thorac Cardiovasc Surg. 1986; 91: 71-78Abstract Full Text PDF PubMed Google Scholar In an animal model, there was a 70% incidence of paraplegia with ligation of the ARM.21Wadouh F Lindemann EF Arndt CM et al.The arteria radicularis magna anterior as a decisive factor influencing spinal cord damage during occlusion.J Thorac Cardiovasc Surg. 1984; 88: 1-10PubMed Google Scholar The combination of the variable origin of the feeding vessels, the segmental nature of the blood supply to the anterior spinal artery, and the discontinuous blood supply of the spinal cord by the anterior spinal artery, accounts for the failure to reliably predict the frequency of SCI after operations on the descending aorta. Occlusion of the aorta results in distal hypotension with a resultant decrease in the perfusion pressure of the spinal cord. In a baboon model, Svensson and Loop3Svensson LG Loop FD Prevention of spinal cord ischemia in aortic surgery.in: Bergan JJ Yao JST Arterial surgery: new diagnostic and operative techniques. Grune & Stratton, New York1988: 273-285Google Scholar have demonstrated a decrease from 20 mL/100 g tissue/min to 1.8 mL/100 g tissue/min in lumbar and thoracic spinal cord blood flow after aortic cross-clamping for 60 min. With 60 min of spinal cord blood flow at 10 mL/100 g tissue/min no paraplegia was observed. On the other hand, when spinal cord blood flow was reduced to 4 mL/100 g tissue/min, all animals became paraplegic. This and many other studies have shown a requirement for a minimum amount of blood flow to prevent paraplegia. Irrespective of the mechanism, a decrease in spinal cord blood flow results in neural ischemia. At normothermia mitochondrial oxidative phosphorylation stops after 3 to 4 min resulting in depletion of the adenosine triphosphate (ATP) stores, and failure of the ATP-dependent membrane pumps, which regulate intracellular calcium homeostasis.22Krause GS White BC Aust SD et al.Brain cell death following ischemia and reperfusion: a proposed biochemical sequence.Crit Care Med. 1988; 16: 714-726Crossref PubMed Scopus (135) Google Scholar The increasing levels of intracellular calcium activate the release of cytoplasmic enzymes that damage DNA and structural proteins. Furthermore, increasing intracellular calcium results in production of xanthine oxidase, which mediates free radical production during reperfusion and release of neurotoxic amino acids aspartate and glutamate.23Rothman S Synaptic release of excitatory amino acid neurotransmitter mediates anoxic neuronal death.J Neurosci. 1984; 4: 1884-1891Crossref PubMed Google Scholar With reperfusion, xanthine oxidase in the presence of reduced nicotinamide adenine dinucleotide phosphate converts molecular oxygen to the superoxide radical. The free radical damage DNA, degrade cellular structural elements, cause a loss of membrane integrity, and increase the ratio of vasoactive prostaglandins. The production of prostaglandins further compounds the ischemia by resulting in vasospasm and microvascular thrombosis. The blood flow to the spinal cord during aortic occlusion is supplied by vessels proximal to the cross-clamp. With aortic occlusion, the aortic pressure above the clamp rises. The proximal aortic hypertension results in an increase in cardiac afterload and an increase in CSF pressure. The risk of paraplegia during aortic surgery is determined by the interaction of four independent processes: (1) decrease in spinal cord blood flow; (2) rate of neuronal metabolism; (3) postischemia reperfusion injury; and (4) postreperfusion blood flow. A number of techniques have been proposed to decrease the risk of SCI after procedures on the descending aorta: expeditious surgery, preservation of intercostal arteries, distal aortic perfusion, drainage of CSF, pharmacological agents, metabolic agents, identification of vascular supply, somatosensory evoked potentials, motor evoked potentials, and hypothermia. These techniques address one or a combination of the determinants of paraplegia. In 1910, Alexis Carrel established a safe aortic cross-clamp time to be between 10 to 15 min during experimental surgery on the aorta.24Carrel A On the experimental surgery of the thoracic aorta and the heart.Ann Surg. 1910; 52: 83Crossref PubMed Google Scholar Clinically the "clamp and go" technique was the first method used in the repair of aneurysms of the thoracic and thoracoabdominal aorta. Livesay et al14Livesay JL Cooley DA Ventemiglia RA et al.Surgical experience in descending thoracic aneurysmectomy with and without adjuncts to avoid ischemia.Ann Thorac Surg. 1995; 39: 37-46Abstract Full Text PDF Scopus (240) Google Scholar showed the risk of paraplegia to increase from 3 to 11% with cross-clamp times higher than 30 min. Svensson and colleagues7Svensson LG Crawford ES Hess KR et al.Experience with 1509 patients undergoing thoracoabdominal aortic operations.J Vasc Surg. 1993; 17: 357-370Abstract Full Text Full Text PDF PubMed Scopus (1084) Google Scholar also determined the aortic cross clamp time to be an independent predictor of SCI. Other studies, however, have failed to show a clear relationship between cross-clamp times and SCI.25Hollier LH Symmonds JB Pairolero PC et al.Thoracoabdominal aortic aneurysm repair: analysis of postoperative morbidity.Arch Surg. 1988; 123: 871-875Crossref PubMed Scopus (129) Google Scholar, 26Najafi H Descending aortic aneurysmectomy without adjuncts to avoid ischemia.Ann Thorac Surg. 1993; 55: 1042-1045Abstract Full Text PDF PubMed Scopus (22) Google Scholar, 27Culliford AT Ayvaliotis B Shemin R et al.Aneurysms of the descending aorta.J Thorac Cardiovasc Surg. 1983; 85: 98Abstract Full Text PDF PubMed Google Scholar Hollier et al25Hollier LH Symmonds JB Pairolero PC et al.Thoracoabdominal aortic aneurysm repair: analysis of postoperative morbidity.Arch Surg. 1988; 123: 871-875Crossref PubMed Scopus (129) Google Scholar studied five patients with paraplegia who had cross-clamp times less than 20 min; conversely, Najafi26Najafi H Descending aortic aneurysmectomy without adjuncts to avoid ischemia.Ann Thorac Surg. 1993; 55: 1042-1045Abstract Full Text PDF PubMed Scopus (22) Google Scholar reported no SCI with a mean cross-clamp time of 58 min. The discrepancy in the results of these studies seems to stem from a failure to delineate the location of the aneurysm and its relationship to the ARM. Katz and colleagues28Katz NM Blackstone EH Kirklin JW et al.Incremental risk factors for spinal cord injury following operation for acute traumatic aortic transection.J Thorac Cardiovasc Surg. 1981; 81: 669-674Abstract Full Text PDF PubMed Google Scholar showed a direct and highly significant correlation between the duration of aortic occlusion and the frequency of neurologic deficit. With up to 25 min of aortic cross-clamping the probability of lower extremity neurologic deficit is virtually nonexistent. After 25 to 30 min of cross-clamping, without other adjuvant techniques, however, the probability of paraplegia or paraparesis rises in a linear fashion such that at 50 min of cross-clamping, virtually all patients exhibit lower extremity neurologic deficits. Obviously, repair of extensive aortic aneurysms requires longer periods of aortic cross-clamping, which would result in unacceptably higher frequency of paraplegia. Although the "clamp and go" technique has been used successfully in most cases, a number of studies have confirmed the need for an additional protective measure with aortic cross-clamp times greater than 30 min.7Svensson LG Crawford ES Hess KR et al.Experience with 1509 patients undergoing thoracoabdominal aortic operations.J Vasc Surg. 1993; 17: 357-370Abstract Full Text Full Text PDF PubMed Scopus (1084) Google Scholar, 14Livesay JL Cooley DA Ventemiglia RA et al.Surgical experience in descending thoracic aneurysmectomy with and without adjuncts to avoid ischemia.Ann Thorac Surg. 1995; 39: 37-46Abstract Full Text PDF Scopus (240) Google Scholar, 27Culliford AT Ayvaliotis B Shemin R et al.Aneurysms of the descending aorta.J Thorac Cardiovasc Surg. 1983; 85: 98Abstract Full Text PDF PubMed Google Scholar, 28Katz NM Blackstone EH Kirklin JW et al.Incremental risk factors for spinal cord injury following operation for acute traumatic aortic transection.J Thorac Cardiovasc Surg. 1981; 81: 669-674Abstract Full Text PDF PubMed Google Scholar Katz et al28Katz NM Blackstone EH Kirklin JW et al.Incremental risk factors for spinal cord injury following operation for acute traumatic aortic transection.J Thorac Cardiovasc Surg. 1981; 81: 669-674Abstract Full Text PDF PubMed Google Scholar reported a 71% incidence of SCI in patients with disease of the descending thoracic aorta and cross clamp times greater than 30 min. Jex et al29Jex RK Schaff HV Piehler JM et al.Early and late results following repair of dissections of the descending thoracic aorta.J Vasc Surg. 1986; 3: 226Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar showed the risk of SCI to decrease from 44 to 8% when distal aortic perfusion was used. In patients undergoing operative repair of dissections of the descending thoracic aorta. A number of techniques have been advocated for perfusing the distal aorta in a retrograde fashion.30Ergin MA Galla JD Lansman SL et al.Distal perfusion methods for surgery of the descending aorta.Sem Thorac Cardiovasc Surg. 1991; 3: 293-299PubMed Google Scholar The fundamental premise has been that increasing the distal aortic perfusion pressure will result in increased blood flow to the entire spinal cord, and thus decrease SCI during aortic cross-clamping. These methods have included roller pumps, passive shunts, and centrifugal pumps. From a historical perspective, roller pumps were used for distal aortic perfusion during the 1960s and 1970s. Partial or total cardiopulmonary bypass can be used to increase distal perfusion beyond the aortic clamps. The various bypass techniques can be used to control the patients' hemodynamics. Unfortunately, the use of a roller pump requires systemic heparinization. In a series of patients undergoing cardiopulmonary bypass for dissection of the aorta from the Massachusetts General Hospital, Stanford, and the Mayo Clinic, 20 to 30% of the patients had complications related to hemorrhage and one third of the deaths were the result of bleeding.31Blaisdell FW Cooley DA The mechanism of paraplegia after temporary thoracic aortic occlusion and its relationship to spinal fluid pressure.Surgery. 1962; 51: 351-355PubMed Google Scholar, 32Svensson LG Grum DF Bednarki M et al.Appraisal of cerebrospinal fluid alterations during aortic surgery with intrathecal papaverine administration and cerebrospinal fluid drainage.J Vasc Surg. 1990; 11: 423-429Abstract Full Text Full Text PDF PubMed Scopus (71) Google Scholar, 33Rothman S Synaptic release of excitatory amino acid neurotransmitter mediates anoxic neuronal death.J Neurosci. 1984; 4: 1884-1891Crossref PubMed Google Scholar, 34Braughler JM Hall ED Central nervous system trauma and stroke I Biochemical considerations for oxygen radical formation and lipid peroxidation.Free Rad Biol Med. 1989; 6: 289-301Crossref PubMed Scopus (604) Google Scholar Some of the earliest resections of the thoracic aortic aneurysms were performed with either temporary internal shunts or a temporary external conduit. The Gott shunt, a unitized polyvinyl chloride catheter with a transparent tridodecyl methylammonium chloride heparin coating was introduced in 1966 and further refined in 1969.35Gott VF Whiffen JD Dutton RC Heparin bonding on colloidal graphite surfaces.Science. 1963; 142: 1297Crossref PubMed Scopus (265) Google Scholar The shunt is placed between the proximal aorta and the segment of the aorta distal to the segment isolated by the cross-clamps. In 1988, Verdant et al36Verdant A Page A Cossette R et al.Surgery of the descending thoracic aorta: spinal cord protection with the Gott shunt.Ann Thorac Surg. 1988; 46: 147-154Abstract Full Text PDF PubMed Scopus (75) Google Scholar reported the use of the Gott shunt in 173 patients undergoing repair of descending thoracic aneurysms. With a mean aortic cross-clamp time of 37 min no paraplegia or other SCI was seen. As a result of the heparin coating, the use of the Gott shunt obviates the need for systemic heparinization. However, in two recently published reports of traumatic tears of the thoracic aorta, the use of the Gott shunt did not decrease the incidence of paraplegia compared with the "clamp and go" technique.9Duhaylongsod FG Glower DD Wolfe WG Acute traumatic aortic aneurysm: the Duke experience from 1970 to 1990.J Vasc Surg. 1992; 15: 331-343Abstract Full Text Full Text PDF PubMed Scopus (85) Google Scholar, 11Hilgenberg AD Logan DL Akins CW et al.Blunt injuries of the thoracic aorta.Ann Thorac Surg. 1992; 53: 233-239Abstract Full Text PDF PubMed Scopus (87) Google Scholar It has been emphasized that to minimize SCI the distal aortic perfusion pressure needs to be greater or equal to 60 mm Hg. Although the flow through a Gott shunt can be monitored with a flow probe, by nature of being a passive shunt the flow cannot be regulated. The inability to regulate the flow through the shunt, results in the inability to insure a distal aortic perfusion pressure of greater or equal to 60 mm Hg. This difficulty makes the Gott shunt less desirable than techniques where the flow can be actively maintained. The inability to regulate flow through the shunt, and furthermore, the inability to retrieve spilled blood make the Gott shunt less desirable in the setting of a thoracic aortic aneurysm in danger of rupture. Centrifugal pumps (Biomedicus pump, Medtronic-Biomedicus Inc; Eden Prairie, Minn) obviate the difficulties associated with active roller pumps as well as passive shunts. These pumps represent the best means of maintaining distal aortic perfusion. Cannulation can be from the left atrium, the pulmonary veins, or the ascending aorta to the descending aorta or the femoral artery. With centrifugal pumps flow can be regulated, damage to the blood components is less than with roller pumps, minimal heparinization is required, and a reservoir can be used in the setting of trauma or a ruptured aneurysm. In an analysis of 596 patients undergoing traumatic rupture of the aorta reported in the English literature, Svensson and Loop3Svensson LG Loop FD Prevention of spinal cord ischemia in aortic surgery.in: Bergan JJ Yao JST Arterial surgery: new diagnostic and operative techniques. Grune & Stratton, New York1988: 273-285Google Scholar calculated the risk of paraplegia to be 2.2%, 2.3%, and 5.8% with bypass, shunts or "clamp and go" techniques, respectively. In these same groups the patients with bypass had a significantly higher mortality rate, 16.7% vs 11.4%, and 5.8% when compared with shunts, and "clamp and go," respectively. Furthermore, they reported a higher incidence of neurologic events with the use of bypass or shunts. This is presumably the result of atheroemboli or air emboli resulting from the cannulation of the diseased aorta.3Svensson LG Loop FD Prevention of spinal cord ischemia in aortic surgery.in: Bergan JJ Yao JST Arterial surgery: new diagnostic and operative techniques. Grune & Stratton, New York1988: 273-285Google Scholar On the other hand, Kaplan et al37Kaplan DK Atsumi N D'Ambra MN et al.Distal circulatory support for thoracic aortic operations: effects on intracranial pressure.Ann Thorac Surg. 1995; 59: 448-452Abstract Full Text PDF PubMed Scopus (16) Google Scholar found active distal bypass techniques to result in maintenance of intracranial pressure, and a significantly greater distal aortic pressure than that achieved with either aortic cross-clamping alone or passive shunting. These authors have reasoned that active distal circulatory support produces the most favorable effect on the two determinants of spinal cord perfusion, intracranial pressure, and distal aortic pressure. Unfortunately the use of distal perfusion techniques does not protect the spinal cord in all patients. Perfusion techniques simply provide blood flow to the distal aorta. If the arteries supplying the anterior spinal artery arise from the excluded segment of the aorta, the spinal cord will remain ischemic even in the face of excellent distal aortic perfusion. The distal perfusion techniques reduce cardiac afterload but also decrease the pressure in the proximal aorta. By reducing the proximal aortic pressure, the perfusion pressure of the spinal cord is decreased. Potentially, this can contribute to further spinal cord ischemia. Furthermore, it has been shown that even when the ARM is perfused using distal perfusion techniques, an increase in blood flow is seen only to the segment of the spinal cord below the junction of the ARM with the anterior spinal artery. This is accounted for by the observation that there is a size discrepancy between the anterior spinal artery above and below the junction of the ARM with the anterior spinal artery. The anterior spinal artery is smaller above the junction with the ARM than below; consequently, because of a decreased resistance, the greater part of the flow is directed to the spinal cord below the junction of the ARM with the anterior spinal artery. Thus even with excellent distal perfusion, the thoracic spinal cord that is fed by the anterior spinal artery above its junction with the ARM remains relatively ischemic.3Svensson LG Loop FD Prevention of spinal cord ischemia in aortic surgery.in: Bergan JJ Yao JST Arterial surgery: new diagnostic and operative techniques. Grune & Stratton, New York1988: 273-285Google Scholar, 38Laschinger JC Spinal cord injury following surgical correction of acute aortic disruption.Sem Thorac Cardiovasc Surg. 1992; 4: 217-222PubMed Google Scholar It has been reasoned that an effective means of preventing paraplegia after surgical procedures on the thoracic aorta is to incorporate the intercostal arteries into the graft. The results of blindly incorporating a large cluster of back-bleeding intercostal arteries between T9 and T12 have been disappointing.39Crawford ES Coselli JS Safi HJ Partial cardiopulmonary bypass, hypothermic circulatory arrest, and posterolateral exposure for thoracic aortic aneurysm operation.J Thorac Cardiovasc Surg. 1987; 94: 824Abstract Full Text PDF PubMed Google Scholar The difficulty with this approach has been that the time involved in reimplanting the intercostals increases the ischemia time, and if the surgeon has guessed incorrectly, this approach leads to a higher risk of paraplegia. A logical approach, therefore, would be to identify preoperatively the vessels that provide the blood flow to the anterior spinal artery. This approach, however, has not gained widespread use because of the fear that direct injection of contrast material into the ARM will result in paraplegia.40Fereshatian A Kadir S Kaufman SL et al.Digital subtraction spinal cord angiography in patients undergoing thoracic aneurysm surgery.Cardiovasc Intervent Radiol. 1989; 12: 7-9Crossref PubMed Scopus (31) Google Scholar, 41Szilagi DE Hageman JH Smith RF et al.Spinal cord damage in surgery of the abdominal aorta.Surgery. 1978; 83: 38-56PubMed Google Scholar, 42DiChio G Doppman J Ommaya AK Selective arteriography of the arteriovenous aneurysms of the spinal cord.Radiology. 1967; 88: 1065-1077Crossref PubMed Scopus (117) Google Scholar The risk of paraplegia in this situation has been attributed to the toxic contrast materials injected at high volumes. In fact, no neurologic consequences have been seen with direct injection of less toxic contrast material into the intercostal arteries.41Szilagi DE Hageman JH Smith RF et al.Spinal cord damage in surgery of the abdominal aorta.Surgery. 1978; 83: 38-56PubMed Google Scholar,43Doppman JL DiChio G Ommaya AK Selective arteriography of the spinal cord. Warren H. Green, Inc, St. Louis1969Google Scholar, 44Kieffer E Richard T Chias J et al.Preoperative spinal cord artriography in aneurysmal disease of the descending thoracic and thoracoabdominal aorta: preliminary results in 45 patients.Ann Vasc Surg. 1987; : 34-46Google Scholar, 45Richard T Guilmet D Bical O et al.La protection medullaire au cours de la chirurgie de l'aorte thoracique et thoracoabdominale.in: Kieffer E Chirurgie de l'aorte thoracique descendante et thoraco-abdominale. L'Expansion Seientifique, Paris, France1986: 104-110Google Scholar Kieffer et al44Kieffer E Richard T Chias J et al.Preoperative spinal cord artriography in aneurysmal disease of the descending thoracic and thoracoabdominal aorta: preliminary results in 45 patients.Ann Vasc Surg. 1987; : 34-46Google Scholar reported a series of 45 patients in whom an attempt was made angiographically to identify the ARM. In 85% of the patients, the ARM was identified on the basis of preoperative contrast studies. In these patients the risk of paraplegia was 5% if the ARM was identified in the excluded segment and that segment of the aorta was completely revascularized. On the other hand, the risk of paraplegia increased to 50% if the segment of aorta bearing the ARM was not revascularized. Overall 60% of the patients in whom the origin of the ARM could not be determined by preoperative angiography became paraplegic. In the same group of patients, the risk of paraplegia was higher when the origin of the ARM was occluded and the vessel filled by collaterals, and when the aorta was clamped for more than 45 min. Williams et al46Wi