Intraoperative Detection of Spinal Cord Ischemia Using Somatosensory Cortical Evoked Potentials during Thoracic Aortic Occlusion

Intraoperative Detection of Spinal Cord Ischemia Using Somatosensory Cortical Evoked Potentials during Thoracic Aortic Occlusion John G. Coles, M.D., Gregory J. Wilson, M.D., Anders F. Sima, M.D., Petr Klement, D.V.M., and Gordon A. Tait, Ph.D.

ABSTRACT Paraplegia remains a devastating and unpredictable complication of surgical procedures requiring temporary occlusion of the thoracic aorta, interruption of important spinal radicular vessels, or both. Intraoperative monitoring of the physiological integrity of the spinal cord should permit the early detection of spinal cord ischemia, the judicious and timely institution of corrective measures, including bypass or shunting, and the preservation of important intercostal arteries in appropriate circumstances. A model of spinal cord ischemia was created by temporary proximal and distal occlusion of the canine thoracic aorta. Serial measurement of somatosensory cortical evoked potentials (SCEP) generated by peripheral nerve stimulation, reflecting the status of long-tract neural conduction, was used to monitor alterations in spinal cord function during ischemia. Twelve animals subjected to aortic occlusion demonstrated a characteristic time-related deterioration of the SCEP with virtual extinction of the signal at a mean interval (+ standard error of the mean) of 12.4 + 1.5 minutes. Six animals in which reperfusion was established immediately following the loss of the SCEP (Group 1) demonstrated complete recovery without neurological sequelae, as assessed by clinical and histological criteria. In 6 animals (Group 2), the period of aortic occlusion was extended for an additional 15 minutes following loss of the SCEP (27.3 f 2.3 minutes); postoperatively, 4 of 6 animals sustained major neurological lesions characterized by spastic paraplegia and histological evidence of spinal cord infarction (Group 1 versus Group 2, p < 0.05).

From The Cardiovascular Laboratories, Banting Institute, University of Toronto, Toronto, Ont, Canada.

Presented at the Eighteenth Annual Meeting of The Society of Thoracic Surgeons, Jan 11-13, 1982, New Orleans, LA.

Address reprint requests to Dr. Wilson, Cardiovascular Laboratories, Room 68, Banting Institute, 100 College St, Toronto, Ont, Canada M5G 1L5.

We conclude that distinctive alterations in the SCEP are indicative of reversible ischemic spinal cord dysfunction. On-line monitoring of spinal cord function using the technique of SCEP provides a rational basis for determining operative strategy during surgical procedures on the thoracic aorta.

Perioperative spinal cord infarction continues to occur with disturbing frequency following resection of thoracic aortic aneurysms. The re- ported incidence of paraplegia complicating the surgical repair of acute traumatic injury to the thoracic aorta ranges from 0 to 23% [l-61. Moreover, ischemic spinal cord injury remains a major hazard following any procedure involving temporary aortic occlusion or the interruption of important radicular vessels, or both [7, 81.

The intraoperative recognition of spinal cord ischemia during periods of aortic occlusion would identify the need for specific technical adjuncts designed to circumvent the development of irreversible neurological injury. The technique of measuring the cortical evoked response generated by peripheral nerve stimulation using signal enhancement techniques has been successfully applied clinically as a means of monitoring the functional integrity of the spinal cord during orthopedic procedures involving mechanical distraction of the spine. In the present study, we investigated the concept that measurement of the somatosensory cortical evoked potential (SCEP), reflecting the fidelity of neural transmission in the spinal cord, could be exploited as a means of detecting ischemic spinal cord dysfunction during periods of aortic occlusion.

With regard to the clinical utility of evoked response monitoring in the context of surgical procedures on the thoracic aorta, it would be critically important to determine if distinctive alterations in this electrophysiological prop-

299 0003-49751821090299-08$01.25 @ 1982 by The Society of Thoracic Surgeons

300 The Annals of Thoracic Surgery Vol34 No 3 September 1982

erty induced by spinal cord ischemia precede the development of irreversible neurological injury. Under these circumstances, direct as- sessment of the functional integrity of the spinal cord provided by intraoperative SCEP monitoring should allow the opportunity for early detection and reversal of ischemic spinal cord injury during surgical procedures on the thoracic aorta.

Material and Methods Experimental Preparation Experiments were performed on 12 adult mon- grel dogs (22 to 30 kg). Following induction with sodium pentobarbital (25 mg per kilogram of body weight) and endotracheal intubation, anesthesia was maintained with enflurane (0.75 to 1.0%). Under sterile operating conditions, a left sixth interspace skin incision with a double intercostal incision through the fourth and eighth intercostal spaces provided access to both the proximal and distal regions of the descending thoracic aorta.

The induction of spinal cord ischemia was accomplished by temporary occlusion of the proximal thoracic aorta, the left subclavian artery, and the distal thoracic aorta at the level of the diaphragmatic hiatus (Fig 1). This method of aortic occlusion creates a reduction in spinal cord blood flow which consistently results in spinal cord infarction if maintained for a 30- minute period at normothermia [9]. Heparin sulfate (1 mglkg) was administered prior to aortic occlusion and was reversed with protamine sulfate (1 mglkg) following release of the vascular clamps.

Recording of t h e SCEP SCEPs were generated by electrical stimulation of the sciatic nerve and recorded with elec- troencephalographic (EEG) scalp electrodes po- sitioned over the primary somaesthetic projec- tion of the contralateral hemisphere (Fig 2). Following exposure of the sciatic nerve in the left popliteal fossa, bipolar stimulation* was performed using two J-shaped electrodes applied directly around the nerve at an inter-

*Grass Model SD, stimulator, Grass Instruments, Quincy, MA 02169

Fig 1 . Method used to accomplish temporary aortic occlusion, showing placement of the vascular clamps on the proximal descending thoracic aorta, the left subclavian artery, and the distal thoracic aorta at the level of the diaphragmatic hiatus.

polar distance of 1 cm. Repetitive stimuli (N =

100 to 150) using a 10 V square-wave stimulus of 5 msec duration at a frequency of 1.2 Hz elicited a consistent pattern of SCEPs.

The evoked electrical potentials were recorded with a single parasagittal vertex elec- trode referenced to the inion. The signal was led to a single-ended, Ac-coupled EEG pre- amplifier,* which was interfaced with a com- puterized signal averager? The latter unit further amplified and filtered the signal using a bandpass filter nominally set at 1 to 3,000 Hz prior to processing. The digitizing rate of the signal averager permitted acquisition of 256 data points during the poststimulus analysis period of 400 to 500 msec corresponding to a temporal resolution of 1.56 to 1.95 msec dwell time per point. A poststimulus analysis time of 400 to 500 msec allowed recording of both the short-latency (primary) and long-latency (secondary and tertiary) components of the cortical evoked response.

The signal averaging process permitted ex- traction of the low-level signal from the back- ground EEG and random electrical noise by

*Grass Model 79 amplifier. tNicolet Model CA-1000 signal averager, Nicolet Biomedical Instruments, Madison, WI 53711.

301 Coles et al: Intraoperative Detection of Spinal Cord Ischemia

Pulse generator

Stimulating electrodes

Fig 2 . The technique used for recording somatosensory cortical evoked potentials. (See text for explanation.)

summing repeated sweeps of data that were time-locked to the stimulus pulse. A trigger signal from the pulse generator, initiated with each pulse, synchronized acquisition of the SCEP by the averaging computer. The compos- ite signal was permanently recorded from the oscilloscopic display of the signal averager by means of a Polaroid camera. The peak-to-peak amplitude and latency of the primary and secondary components of the evoked response were measured with a calibrated cursor on the computer oscilloscope or directly from the photographed recording.

Experimental Protocol Superimposition of two independently collected averages confirmed the integrity of the SCEP prior to aortic occlusion in each experiment. Measurement of the SCEP was repeated at 3-minute intervals during the period of aortic occlusion beginning 1 minute following the onset of occlusion, and at 5, 15, and 30 minutes following release of the clamps (reperfusion). Approximately 2 to 2.5 minutes, measured from the start of each time interval, were required to generate each summated recording.

The duration of aortic occlusion was determined in each experiment by the results from on-line recording of the SCEP. In Group 1 (6 animals), aortic occlusion was maintained until substantial attenuation was evident in both the primary and secondary components of the evoked response, after which reperfusion was

immediately established. For purposes of analysis, a 50% or greater reduction in the peak- to-peak amplitude of the primary or secondary component of the evoked response was consid- ered indicative of substantial attenuation (or loss) of the corresponding signal. In Group 2 (6 animals), the period of aortic occlusion was extended for an additional 15 minutes following 50% or greater attenuation of the SCEP.

Following removal of the vascular clamps and completion of the SCEP measurements, the chest was closed without postoperative tube drainage. All animals were recovered for as- sessment of functional neurological recovery and eventual histological examination of the spinal cord. The status of neurological function, determined at 24 hours and 7 days postoperatively, was designated according to Tarlov’s [lo] criteria for grading the recovery of hind limb motor function: 0 = no voluntary movement; 1 = perceptible movement of joints; 2 = good movements at joints but inability to stand; 3 = ability to stand and walk; and 4 = complete recovery. Sensory testing involving the response to painful stimuli presented to the hind limb was performed at the same postoperative time intervals. All animals were killed 7 days postoperatively, and the thoracic, lumbar, and sacral segments of the spinal cords, including the cauda equina, were explanted. The specimens were sectioned at 2 to 4 cm intervals, prepared for light microscopy, and examined by one of us (A.F.S.), a neuropathologist, without knowledge of the group assignment.

StatisticaZ Analys i s The Student t test for paired data was used in computing the p value for comparison of the ischemic intervals at which loss of the primary and secondary components of the SCEP occurred. Fisher’s exact test was used to detect significant differences in proportions with respect to outcomes between the two groups. The results of continuous observations are ex- pressed as the mean f standard error of the mean (SEM).

Results Figure 3 depicts a typical preocclusion tracing of the SCEP. The characteristic features of the


4 5u v 4

,A . . ..J 50ms

SCEPl SCEP2 - A

CONTROL 4-2

Fig 3. Tracing of the somatosensory cortical evoked potentials (SCEP) recorded in a typical experiment prior to induction of spinal cord ischemia. The SCEP consists of an early surface-positive response (SCEP,) with a peak latency of 20 msec, followed by a secondary surface-positive response (SCEP,) with a peak latency of 65 msec and a series of less conspicuous tertiary waves of longer latency. The dashed line represents a metachronously collected recording confirming the re- liability of the SCEP prior to aortic occlusion. Positivity is represented by upward deflection. The arrow indicates the stimulus artifact.

SCEP consist of an early positive response (SCEP,; peak latency = 18.5 f 1.0 msec), followed by a secondary positive response (SCEP,; peak latency = 79.1 f 8.0 msec) of greater amplitude and duration and a series of long- latency tertiary waves of variable configuration.

Following the induction of spinal cord ischemia, a progressive time-related attenuation of the evoked response was observed in all experiments (Fig 4). Substantial (>50%) diminu- tion in the amplitude of the SCEP, occurred after 4.0 f 1.2 minutes of ischemia and preceded a similar reduction in amplitude of the SCEP2, which occurred at 12.6 f 1.6 minutes ( p < 0.005). Thus, loss of the early component of the evoked response, SCEP,, was the most sen- sitive indicator of ischemic spinal cord dysfunction.

Following substantial loss of both the SCEPl and SCEP2 components of the evoked response, reperfusion was established at 12.4 k 1.5 minutes in Group 1 animals, whereas the ischemic interval was extended for an additional 15- minute period to 27.3 f 2.3 minutes in Group 2 animals. Group 1 animals uniformly recovered, without clinical evidence of neurological injury as assessed at 24 hours and 7 days postoperatively. In contrast, 4 of 6 Group 2 animals sustained a significant perioperative motor lesion (Group 1 versus Group 2, p < 0.05

POSTOCCLUSION lmin

22min

1 REPERFUSION 5min

30min ‘W Loss of SCEPl - lmin Loss of SCEP2 - lOmin

Occlusion time - 25min

Fig 4. Tracings of the somatosensory cortical evoked potentials (SCEP) recorded during a typical experiment, illustrating the time course of the evoked response during ischemia and following reperfusion. Note loss of the SCEP, following 1 minute of ischemia, with substantial attenuation of secondary (SCEP,) and tertiary activity within 10 minutes postocclusion. In this experiment (Group 2), aortic occlusion was continued for an additional 15-minute period, during which time there was virtual electrical silence with no recognizable evoked response recorded. Reperfusion was established following 25 minutes of aortic occlusion. Note substantial recovery of the SCEP, following 30 minutes of reperfusion.

by Fischer’s exact test). Three of the 4 affected animals demonstrated a complete spastic paraplegia, evident at 24 hours with no im- provement at 7 days. The other affected animal demonstrated only grade 2 motor function (good movements at joints but inability to stand) at 24 hours, improving to grade 3 motor function (ability to stand and walk) at 7 days postoperatively.

Although the results of postoperative sen-


sory testing were difficult to quantify, painful stimuli presented to the hind limb evoked a normal withdrawal response in all 6 Group 1 animals as well as the 2 Group 2 animals without motor impairment. Furthermore, a positive response to painful stimuli also was elicited in 2 of 4 animals with complete motor lesions, im- plying a degree of dissociation in the extent of involvement of the motor and sensory pathways. The clinical findings were paralleled by the status of the SCEP, following 30 minutes of reperfusion, which showed substantial recovery (>50°/0 of preocclusion amplitude) in 9 of 10 animals with preservation of sensory function.

The results of the histological evaluation sub- stantiated the clinical findings. The spinal cords from the animals with complete motor lesions demonstrated areas of extensive infarction involving the gray matter in the distal lumbar and proximal sacral segments (Fig 5A). The corresponding white matter revealed more subtle changes consisting of limited demyelination and gliosis affecting principally the ventral and lateral columns. These histological findings sharply contrast with those of Group 1 and the 2 unaffected Group 2 animals, which were lim-

Fig 5. (A) Microscopic section of the distal lumbar segment of spinal cord taken from an animal with severe spastic paraplegia postoperatively. Extensive replace- ment of the central gray matter with macrophages is indicative of previous infarction. Residual degenerating anterior horn cells are present (arrow). ( B ) Corre- sponding region of spinal cord taken from an animal without clinical evidence of neurological injury. Note the preservation of normal-appearing anterior horn cells and absence of the macrophage reaction. (Both Ha€; xl00 before 40% reduction.)

ited to a minimal degree of neuron loss and gliosis involving the central gray matter of the lumbosacral cord (Fig 5B).

Comment The prevention of ischemic injury to the spinal cord during repair of lesions of the descending thoracic aorta remains a challenging surgical problem. Temporary occlusion of the thoracic aorta, in the absence of adequate collateral vessels or effective bypass, may compromise spinal cord blood flow indirectly as a result of a critical reduction in distal aortic perfusion. The risk of perioperative spinal cord infarction is increased by the necessity for concomitant resection or exclusion of spinal radicular vessels, irrespec- tive of the adequacy of distal aortic perfusion. Furthermore, inability to localize the origin of important radicular vessels in individual patients precludes the use of a systematic operative strategy that will uniformly prevent a critical reduction in spinal cord perfusion during periods of aortic occlusion.

Intraoperative recognition and reversal of spinal cord ischemia has not been successfully accomplished during the performance of surgical procedures on the thoracic aorta. The results of the present study suggest that serial measurement of SCEPs, indicative of the status of conductivity in the neuraxis, could be exploited as a means of detecting ischemic spinal cord dysfunction during occlusion of the thoracic aorta. A characteristic time-related deterioration in the evoked response was observed following the induction of spinal cord ischemia, with virtual cessation of neural transmission


within a 12-minute period. This finding is con- sonant with that of Kobrine and colleagues, who measured the evoked response in monkeys using direct dural recordings from the cer- vicomedullary region [ l l l and the somatosensory cortex [12]. These investigators noted the onset of electrical silence following 8 to 18 minutes of profound ischemia induced by con- trolled exsanguination to below the limits of circulatory autoregulation in the spinal cord.

The available evidence indicates that the earliest surface-positive potential (SCEP,) in animals is generated by the ascending afferent volley in the lemnisco-thalamocortical pathways and the contralateral primary sensory cortex, and is analogous to the more complex short-latency cortical evoked response observed in recordings done in humans [131. The precise neuroanatomical substrate of the secondary component (SCEP,) and subsequent long- latency components of the animal (and human) evoked response, characterized by diffuse and symmetrical cortical radiation, remain to be completely delineated [14, 151.

In the present study, a consistent finding was the rapid loss of the earliest surface-positive potential (SCEP,) following induction of spinal cord ischemia, which preceded eventual loss of the secondary and tertiary components of the evoked response. The reason for the more rapid loss of the initial component of the evoked response, which represents the earliest indication of ischemic spinal cord dysfunction, is un- known. Selective loss of the initial deflection of the cortical evoked response has been recorded in patients with documented vascular lesions involving the thalamus [161. It may be specu- lated that altered spinal cord conduction during ischemia selectively increases the latency of the SCEP,, which then becomes buried, as a result of temporal dispersion, in the more conspicuous secondary component of the evoked response.

Intraoperative monitoring of spinal cord function with SCEPs has been used successfully in the detection of injurious mechanical distraction of the spinal cord during the correction of scoliosis by Harrington instrumentation [17, 181, and during decompression and stabilization

procedures following traumatic lesions of the thoracic spine [19]. Our results indicate that distinctive alterations in the SCEP are associated with ischemic dysfunction of the spinal cord occasioned by temporary occlusion of the thoracic aorta. Spinal cord ischemic injury of sufficent degree to result in abolition of the cortical evoked response, if promptly followed by restoration of spinal cord perfusion, was uniformly compatible with complete recovery of neurological function. Thus, the early recognition of spinal cord ischemia using on-line measurement of SCEP allows the opportunity for rational intervention with corrective measures designed to enhance spinal cord perfusion during surgical procedures on the thoracic aorta.

In this context, the development of characteristic changes in the SCEP following cross- clamping of the thoracic aorta would indicate the need for adjunctive measures, such as bypass or a shunt, to improve perfusion distal to the cross-clamped aortic segment and, indirectly, to the spinal cord. Recovery of the SCEP following this intervention would indicate that the intended procedure could be completed without the risk of perioperative spinal cord infarction. Conversely, failure to recover the normal pattern of the SCEP would constitute strong circumstantial evidence that spinal cord blood flow is inadequate because an important spinal radicular vessel, typically the arteria radicularis magna [20, 211, has been excluded from systemic perfusion as a result of the posi- tion of the vascular clamps. Given this exi- gency, and lacking specific information relating to the level of origin of the arteria radicularis magna, reimplantation of a segment of posterior aortic wall from which several intercostal arteries originate would be imperative in order to prevent irreversible spinal cord injury. Available experimental evidence suggests that the efficacy of intercostal reimplantation would be enhanced by the induction of even modest levels of systemic hypothermia [12, 22-24], since the time required to complete such a procedure may exceed the tolerable period of nor- mothermic ischemia of the corresponding segment of the spinal cord [25].


Alternatively, evidence of spinal cord ischemia despite effective distal aortic perfusion may justify abandonment of the intended procedure, given the risks and uncertainty at- tending random reimplantation of intercostal vessels. Angiographic demonstration of the arteria radicularis magna by selective catheteri- zation [26] could then be performed electively as an interval procedure prior to definitive repair of the lesion at a second operation. Under these circumstances, precise knowledge of the level of origin of major spinal arteries would allow reimplantation or preservation of the appropriate intercostal vessels, thus minimizing the risk of perioperative spinal cord infarction.

Theoretically, an ischemic lesion of the spinal cord could selectively involve the motor neurons in the gray matter without apparent perturbation in the SCEP, which is principally mediated through posterior column pathways located in the white matter 1271. In the present study, this dissociative type of injury was evident in 2 animals subjected to extended aortic occlusion following loss of the evoked response; they sustained major motor lesions in association with preservation of sensory function and substantial recovery of the reperfusion evoked response. This finding suggests that a reduction in spinal cord blood flow of sufficient severity to produce isolated infarction of motor pathways is associated with at least temporary ischemic dysfunction of the somatosensory pathways and corresponding abnormalities in the evoked response.

These results demonstrate that the character- istics of the SCEP can be profitably analyzed to provide useful information regarding the functional status of the spinal somatosensory sys- tem under conditions of aortic occlusion. We conclude that the concept of spinal cord monitoring may assume an important role in the prevention of ischemic injury to the spinal cord during surgical procedures on the thoracic aorta.

References 1. Turney SZ, Attar S, Ayella R, et al: Traumatic

rupture of the aorta. J Thorac Cardiovasc Surg 72:727, 1976

2. Stallone RJ: Management of ruptured thoracic aortic aneurysms. Am J Surg 136:162, 1978

3. Kirsh MM, Behrendt DM, Orringer MB, et al: The treatment of acute traumatic rupture of the aorta. Ann Surg 184:308, 1976

4. May IA, Ecker RR, Iverson LG: Heparinless femoral venoarterial bypass without an oxygen- ator for surgery on the descending thoracic aorta. J Thorac Cardiovasc Surg 73:387, 1977

5. Wakabayashi A, Connolly JE: Prevention of paraplegia associated with resection of extensive thoracic aneurysms. Arch Surg 111:1186, 1976

6. Katz NM, Blackstone EH, Kirklin JW, Karp RB: Incremental risk factors for spinal cord injury following operation for acute traumatic aortic transection. J Thorac Cardiovasc Surg 81:669, 1981

7. Crawford ES, Synder DM, Cho GC, Roehm JF: Progress in treatment of thoracoabdominal and abdominal aortic aneurysms involving coeliac, superior mesenteric, and renal arteries. Ann Surg 188:505, 1978

8. Adams HF: Neurologic complications of aortic surgery. Ann Surg 144:574, 1956

9. Coles JG, Wilson GJ, Klement P, et al: Intraoper- ative management of thoracic aortic aneurysm: experimental evaluation of perfusion cooling in the prevention of ischemic injury to the spinal cord (abstract). J Thorac Cardiovasc Surg (in press)

10. Tarlov IM: Spinal Cord Compression: Mecha- nism of Paralysis and Treatment. Springfield, IL, Thomas, 1957, p 147

11. Kobrine AI, Evans DE, Rizzoli HV: The effects of ischemia on long-tract neural conduction in the spinal cord. J Neurosurg 50:639, 1979

12. Kobrine AI, Evans DE, Rizzoli HV: Relative vul- nerability of the brain and spinal cord to ischemia. J Neurol Sci 45:65, 1980

13. Allison T, Goff WR, Williamson I'D, van Gilder JC: On the neural origin of early components of the human somatosensory evoked potential. In Desmedt JE (ed): Clinical Uses of Cerebral, Brainstem and Spinal Somatosensory Evoked Potentials (Progress in Clinical Neurophysiol- ogy, Vol 7). Basel, Karger, 1980, pp 51-68

14. Goff WR, Williamson I'D, van Gilder JC, et al: Neural origins of long latency evoked potentials recorded from the depth and from the cortical surface of the brain in man. In Desmedt JE (ed): Clinical Uses of Cerebral, Brainstem and Spinal Somatosensory Evoked Potentials (Progress in Clinical Neurophysiology, Vol 7). Basel, Karger,

15. Williamson PD, Goff WR, Allison T: Somatosen- sory evoked responses in patients with unilateral cerebral lesions. Electroencephalogr Clin Neuro- physiol 28:566, 1970

1980, pp 126-145


16. Noel P, Desmedt JE: Somatosensory cerebral evoked potentials after vascular lesions of the brain-stem and diencephalon. Brain 98:113, 1975

17. Brown RH, Nash CL: Current status of spinal cord monitoring. Spine 4:466, 1979

18. Nash CL, Lorig RA, Schatzinger LA, Brown RH: Spinal cord monitoring during operative treatment of the spine. Clin Orthop 126:100, 1977

19. Spielrolz NI, Benjamin MV, Engler GL, Ranshoff J: Somatosensory evoked potentials during decompression and stabilization of the spine. Spine 4:500, 1979

20. Adamkiewicz A: Die Blutgefasse des menschli- chen Ruckenmarkesoberflache. Sitz Akad Wiss Wien Math Natur Klass 85:101, 1882

21. Gillilan LA: The arterial blood supply of the human spinal cord. J Comp Neurol 75:110, 1958

22. Pontius RG, Buckman HL, Hardy EG, et al: The use of hypothermia in the prevention of paraplegia following temporary aortic occlusion: experimental observations. Surgery 36:33, 1954

23. Beattie EJ, Adavasic D, Keshishian JM, Blades B: Refrigeration in experimental surgery of the aorta. Surg Gynecol Obstet 96:711, 1953

24. Parkins WS, Ben M, Vars HM: Tolerance of temporary occlusion of the thoracic aorta in nor- mothermic and hypothermic dogs. Surgery 38:38, 1955

25. Szilagyi DE, Hageman JH, Smith RF, Elliot JP: Spinal cord damage in surgery of the abdominal aorta. Surgery 83:38, 1978

26. Doppman JL, Di Chiro G, Morton DL: Arterio- graphic identification of spinal cord blood supply prior to aortic surgery. JAMA 204:172, 1978

27. Dorfman LJ, Perkash I, Bosley Th4, Cummins KL: Use of cerebral evoked potentials to evaluate spinal somatosensory function in patients with traumatic and surgical myelopathies. J Neu- rosurg 52:654, 1980

Discussion DR. JOSEPH N. CUNNINGHAM (New York, NY): I con- gratulate Dr. Coles and his group on a fine study. At New York University we have performed laboratory experimentation very similar to that which Dr. Coles and his group have so convincingly presented today. On the basis of our findings we have routinely applied these techniques directly in the operating room for the past several weeks for all surgery performed on the thoracic aorta.

We now have accumulated 6 consecutive patients who underwent large resections of descending thoracic or thoracoabdominal aneurysms. In 5 of them, somatosensory evoked potentials began to disappear within 3 minutes following application of a proximal cross-clamp. All potentials were completely gone at 9 minutes. This also occurred in the sixth patient, but in the first 5, reversal of the somatosensory evoked potentials could be obtained by use of a shunt or femorofemoral bypass. In the sixth patient, a critical intercostal artery was reimplanted, resulting in immediate return of the somatosensory evoked potentials to normal.

We believe that both from the standpoint of the patient’s welfare with regard to prevention of paraplegia, and with respect to the obvious medicolegal implications, the time has come for continuous monitoring of evoked potentials in the operating room during aortic cross-clamping to become part of routine hospital policy.

DR. COLES: I was very interested in your comments, Dr. Cunningham. At present we have no clinical ex- perience with this technique but would like to con- gratulate you for demonstrating the potential clinical utility of spinal cord monitoring. We hope that our results have provided some experimental validation for the concept of spinal cord monitoring during surgical procedures on the thoracic aorta.

Documents

Intraoperative Detection of Spinal Cord Ischemia Using Somatosensory Cortical Evoked Potentials during Thoracic Aortic Occlusion