Monitoring of somatosensory evoked potentials during surgical procedures on the thoracoabdominal aorta

J

THoRAc CARDIOVASC SURG

1987;94:260-5

Monitoring of somatosensory evoked potentials during surgical procedures on the thoracoabdominal aorta 1. Relationship of aortic cross-clamp duration, changes in somatosensory evoked potentials, and incidence of neurologic dysfunction To determine if intraoperative monitoring of somatosensory evoked potentials detects spinal cord ischemia, we subjected 21 dogs to aortic cross-clamping distal to the left subdavian artery. Group I animals (short-term studies, n = 6) demonstrated decay and loss of somatosensory evoked potentials at 8.5 ± 1.1 minutes after aortic cross-clamping. During loss of somatosensory evoked potentials, significant decreases in spinal cord blood flow occurred in cord segments below T 6" Significant reactive hyperemia occurred without normalization of somatosensory evoked potentials after reperfusion. Fifteen Group iI animals (long-term studies) were studied to determine the relationship between duration of spinal cord ischemia (evoked potential loss) and subsequent incidence of paraplegia. Extension of aortic cross-clamping for 5 minutes after loss of somatosensory evoked potentials in six dogs resulted in no paraplegia (mean cross-clamp time 12.7 ± 0.6 minutes). Prolongation of aortic cross-clamping for 10 minutes after evoked potential loss in nine dogs (mean cross-clamp time 17.6 ± 0.6 minutes) resulted in a 67% (6/9) incidence of paraplegia 7 days postoperatively (p = 0.02 versus 10 minutes of aortic cross-clamping). These findings demonstrate that simple aortic cross-clamping uniformly results in spinal cord ischemia and that such ischemia is detectable by monitoring of somatosensory evoked potentials. Duration of ischemia, as measured by the time of evoked potential loss during the cross-clamp interval, is related to the incidence of postoperative neurologic injury.

John C. Laschinger, M.D., Joseph N. Cunningham, Jr., M.D., Matthew M. Cooper, M.D., F. Gregory Baumann, Ph.D., and Frank C. Spencer, M.D., New York and Brooklyn, N. Y.

RraPlegia and paraparesis are catastrophic complications of operations involving the descending thoracic and thoracoabdominal aorta. Despite numerous surgical and anesthetic approaches':' and the use of temporary shunt or partial bypass techniques.>'? the frequency of these

From the Divisions of Cardiovascular and Thoracic Surgery, Departments of Surgery, New York University Medical Center, New York, N. Y., and Maimonides Medical Center, Brooklyn,

N. Y. Received for publication July 22, 1986. Accepted for publication Aug. 23, 1986. Address for reprints: Joseph N. Cunningham, Jr., M.D., Director, Department of Surgery, Maimonides Medical Center, 4802 l Oth Ave., Brooklyn, N. Y. 11219.

260

complications ranges from 0.5% for operations for coarctation of the aorta" to 38% after operations for extensive thoracoabdominal aneurysms." The potential for severe neurologic damage associated with operations on the thoracoabdominal aorta will remain high until methodology that can detect and guide the surgeon in prevention of intraoperative spinal cord ischemia is routinely employed." The purpose of this experimental study was to determine if intraoperative monitoring of somatosensory evoked potentials (SEPs) would accurately characterize the adequacy of spinal cord blood flow and detect spinal cord ischemia during periods of aortic cross-clamping analogous to those encountered clinically. We sought to define the relationship between duration of spinal cord ischemia (as reflected by changes

Volume 94 Number 2

Monitoring of SEPs, I

AU9ust 1987

26 1

Cortical Response •

200

consecutive

stimuli

Signal Input • •

300V 0.6 msec

•

2.3/sec

Fig. 1. Schematic representation of technique employed for SEP monitoring in experimental studies.

in SEPs) and the subsequent incidence of ischemic neurologic sequelae. Material and methods Twenty-one mongrel dogs were intubated and anesthetized for study. Morphine sulfate (2.5 mg/kg) and alpha chloralose (lOO mg/kg) were used for short-term studies (n = 6) and pentobarbital sodium (Nembutal, 25 mg/kg) for long-term studies (n = 15). Intra-arterial pressure was measured in the ascending aorta and the femoral artery. Radioactive microspheres (iodine 125, cerium 141, strontium 85, and selenium 46) were injected through a left atrial catheter and serial reference samples for determination of spinal cord blood flow were obtained from the ascending aorta." SEP traces were generated with a clinical evoked potential system* by stimulation of the posterior tibial nerves with a bipolar input channel (Fig. 1). Impulses conducted by the dorsal spinal columns were recorded from midline scalp electrodes at the nasion and 55% of the distance from the nasion to the inion. A separate grounding electrode was placed in the left ear. The potentials were amplified 10,OOOX and then processed with a 10 Hz LoPass and a 250 Hz HiPass filter.* To improve the signal-to-noise ratio of these small potentials, 200 consecutive responses activated by supramaximal stimuli to the nerve (4X motor twitch threshold, 20 mAmps, 0.4 msec pulse duration, 2.3/sec) were averaged for each tracing. An additional thoracic grounding electrode was placed and a no-stimulus control trace was recorded for establishment of background noise. A typical SEP trace is shown in Fig. 2. Two parameters were serially monitored: latency of onset and amplitude of the generated response. Repetitive genera*TN-3,OOO Tracor Analytic, Inc., Elk Grove Village, III.

)4V

msec Fig. 2. Typical SEP trace showing parameters of latency of onset and amplitude of response.

tion of new tracings that show increases in latency or decreases in amplitude, or both, indicate spinal cord ischemia. IS, 16 Short-term studies (Group I, n = 6). The thoracic aorta was cross-damped distal to the left subclavian artery for 60 minutes followed by 20 minutes of reperfusion. Mean proximal aortic pressure was maintained at 80 mm Hg with intravenous sodium nitroprusside. Serial measurements of spinal cord blood flow and SEPs were conducted. Animals were killed, a posterior laminectomy performed, and the extirpated spinal cord was cut into sections (CI-T s, T6-13' L I _s) for determination of spinal cord blood flow. Long-term studies (Group II, n = 15). In 15 sterile experiments, the aorta was cross-damped at the same level until SEP loss was noted. In six animals (Group IIa) the period of aortic occlusion was extended for an additional 5 minutes after SEP loss and in nine animals (Group lIb) for an additional 10 minutes after SEP loss before reperfusion. Sequential measurements of SEP and spinal cord blood flow during cross-damping and reperfusion were repeated. Neurologic evaluation as

The Journal of

2 6 2 Laschinger et al.

Thoracic and Cardiovascular Surgery

Table I. Grading system used for evaluation of postoperative neurologic status (adapted from Tarlov 17) Grade Grade Grade Grade Grade Grade

0 I 2 3 4

Table II. Summary of latency and amplitude changes in Group I animals during 60 minutes of aortic cross-elamping

Description No voluntary movement; spastic paraplegia Perceptible movement of joints Good movements at joints but inability to stand Ability to stand and walk Complete recovery

described by Tarlov" and SEP determinations were performed on postoperative days 1, 3, and 7, and the degree of neurologic impairment was graded on a four-point scale (Table I). The surviving animals underwent repeat left thoracotomy and posterior laminectomy on postoperative day 7 for injection of radioactive microspheres and cord extirpation to determine spinal cord blood flow. All animals used in this and the associated experiments reported in this experimental series received humane care in compliance with the "Principles of Laboratory Animal Care" formulated by the National Society for Medical Research and the "Guide for the Care and Use of Laboratory Animals" prepared by the National Academy of Sciences and published by the National Institutes of Health (NIH Publication No. 80-23, revised 1978). Unless otherwise noted, statistical significance of differences was determined by analysis of variance, and all values are reported as mean ± the standard error of the mean Results Short-term experiments. At 3.5 minutes after aortic cross-elamping, significant changes in SEPs were noted (Table II). The latency increased approximately 7% in association with a simultaneous diminution in amplitude of 41 % (p < 0.05 and p < 0.005 versus baseline, respectively, paired Student's t test). These changes were progressive until complete loss of the SEPs was noted at 8.5 minutes after aortic cross-clamping. Aortic reperfusion after the 60 minute ischemic interval was associated with minimal return of SEPs with severe prolongation of latency and marked diminution in amplitude. SEP traces (Fig. 3) constructed from the data in Table II illustrate these significant changes, previously described as a type I SEP response." A striking observation illustrated in Table III was the demonstration of a statistically significant decrease in spinal cord blood flow to areas of the lower thoracic (T6-13) and lumbosacral (L 1-s) spinal cord in association with the loss of SEPs. Lower thoracic blood flow

Latency (msec) Baseline Aortic cross-clamp time 3.5 ± 0.7 min 8.5 ± 1.1 min Reperfusion (20 min)

Amplitude (mV)

21.0

± 1.14

3.44 ± 0.53

22.5

± 1.27*

2.05 ± 0.46*

t

29.68 ± 3.55*

t

0.96 ± 0.30*

'p < 0.05 versus baseline (paired Student's t test). [No electrical activity.

diminished from 4.29 ± 0.90 to 1.89 ± 1.01 ml/IOO gm/rnin, and lumbosacral flow decreased from 6.40 ± 1.34 to 2.21 ± 0.68 ml/IOO gm/rnin after 20 minutes of aortic cross-clamping and total SEP loss. Cervical and upper thoracic flow remained unchanged during this interval, consistent with the known blood supply to these segments. Reperfusion to the distal aorta after 60 minutes of ischemia resulted in a significant reactive hyperemic response in the lower thoracic and lumbosacral segments, most evident 20 minutes after unclamping. Baseline flows in the lower thoracic segments quadrupled (4.29 ± 0.90 versus 17.43 ± 7.27 ml/IOO gm/min) and a sixfold increase in lumbosacral flow was noted (6.40 ± 1.34 versus 38.66 ± 9.48 ml/IOO gm/ min), both statistically significant changes. Concomitantly, no significant change in cervical-upper thoracic cord flow was noted (6.89 ± 1.35 versus 5.14 ± 1.28 ml/loo gm/rnin), Long-term experiments. A significant increase in the incidence of severe neurologic impairment was noted when duration of aortic cross-clamping after complete loss of SEPs extended beyond 5 minutes, Recovery in Group IIa animals was complete with no evidence of neurologic impairment resulting from a mean crossclamp time of 13 minutes (Table IV). Latency and amplitude of the SEP trace in five of six animals returned to baseline during a 20 minute reperfusion interval and remained normal. In contrast, a mean cross-clamp interval of 18 minutes in Group lIb animals resulted in a 67% (6/9) incidence of severe neurologic injury (p = 0.02 versus Group IIa for incidence of neurologic injury grade 0 to 3, Fisher's exact test). One animal had an ataxic gait, and five had complete spastic paraplegia. Clearly, once altered, the return of SEP tracings to normal did not correlate with ultimate neurologic status. In fact, three of five animals having spastic paraplegia regained normal SEP traces at postoperative day 3. In contrast, all animals with normal neurologic status postoperatively exhibited normal SEPs by the first postoperative day.

Volume 94 Number 2 August 1987

Monitoring of SEPs, I

263

BASELINE

Axe

35

Axe

8.5 ± 1.1 min

± 0.7 min \1---....

REPERFUSION 20 min

Fig. 3. Type I SEP response. Note prolongation of latency and diminution of amplitude during progression to complete SEP loss. AXe, Aortic cross-clamping.

Measurements of spinal cord blood flow after complete SEP loss in both groups subjected to long-term experiments revealed similar significant decreases in bloodflow in lower thoracic and lumbosacral spinal cord segments, as was previously noted in the short-term experiments (Tables V and VI). Of importance is the - fact that repeat measurements of spinal cord blood flow obtained 7 days postoperatively in both Group IIa and lIb dogs revealed no significant differences from baseline in either group and no correlation between blood flow and ultimate neurologic status.

Discussion Numerous clinical and experimental investigations have suggested a relationship between the duration of aortic cross-clamp time and the production of paraplegia.!" Paraplegia may occur in some patients after even the briefest of cross-clamp intervals," whereas other patients with increased collateral circulation associated with chronic aortic lesions and arteriosclerotic disease may tolerate much longer ischemic intervals without permanent neurologic damage. I, 2 The results of this experimental study confirm the important relationship between duration of aortic crossclamping and production of postoperative neurologic injury in the noncollateralized canine spinal circulation. With a clinical modality of monitoring evoked spinal cord potentials, it has been possible to demonstrate that ischemiasevere enough to cause spinal cord dysfunction occurs within 3 to 5 minutes after aortic cross-clamping and results in complete loss of spinal cord conduction after 7 to 9 minutes. Prolongation of this ischemia for periods as short as 10 minutes after complete loss of SEPs results in permanent spinal cord injury in a majority (67%) in the canine experimental model. However, if reperfusion is established in a shorter interval of time (less than 5 minutes) after SEP loss

Table m. Changes in spinal cord blood flow (mlllOO gmimin] after aortic cross-clamping in Group I animals Spinal cord segment C,-T, To." L,.s

Baseline

20 min after AXe (SEP loss)

20 min after reperfusion (SEP return)

6.89 ± 1.35 4.29 ± 0.90 6.40 ± 1.34

8.40 ± 2.99 1.89 ± 1.01* 2.21 ± 0.68t

5.14 ± 1.28 17.43 ± 7.27* 38.66 ± 9.48t

Legend: AXe, Aortic cross-clamping. SEP, Somatosensory evoked potential. "p < 0.05 versus baseline. tp < 0.025 versus baseline.

occurs, there appears to be no incidence of permanent ischemic spinal cord injury. Although not totally analogous to the human model, these experimental data suggest that brief periods of spinal cord ischemia may produce significant neurologic injury in the noncollateralized spinal circulation of humans. Proximal cross-clamping of the thoracic aorta distal to the left subclavian artery results in significant diminution in blood flow to the lower spinal cord from the level of T 6 through the lumbosacral segments. After cross-clamping, overall cord flow is reduced at the expense of these lower segments whereas the greater percentage of total cord flow remains in the cervical and upper thoracic distributions. These observations are consistent with anatomic descriptions of spinal cord blood supply wherein upper cord levels (above T 6) are supplied by vessels originating from the vertebral and subclavian arteries and lower cord blood supply arises from segmental branches originating from the aorta at variable levels." Thus, since the lower cord below T 6 is most vulnerable after proximal aortic cross-clamping, it is not surprising that these present experiments demonstrated significant ischemia in these segments followed

The Journal of Thoracic and Cardiovascular Surgery

2 6 4 Laschinger et al.

Table IV. Summary ofpostoperative neurologic status and SEP return in Group IIa and IIb dogs

AXe duration after Dog No.

SEP retumi

SEP loss (min)

Neurologic score*

5 5 5 5 5 5

4 4 4 4 4 4

+ + + +

10 10 10 10 10 10 10 10 10

4 0 0 4 0 3 0 0 4

± ± ± ± ± ±

20 min

Group lIa I 2

3 4 5 6

±

+

Group lIb 1 2 3 4 5 6 7 8 9

+ +

I

1 day

I

3 days

I

7 days

+ + + + + +

+ + + + + +

+ + + + + +

+

+ +

+ +

+

+

+

±

+ + + +

+ + + +

± ±

+ + +

Legend: AXC, Aortic cross-clamp. SEP, Somatosensory evoked potentials. • Final score, day 7. = Return to baseline latency and amplitude; amplitude versus baseline).

t+

= SEP absent;

±

= SEP present

but abnormal compared to baseline (> 10% change in latency and/or 40%change in

Table V. Group IIa: Changes in spinal cord blood flow (ml/l00 gmfmin] in animals exhibiting 5 minutes of complete SEP loss (AXC time = 13 minutes) Spinal cord segment

Baseline

SEP loss

20 min after reperfusion

7 days postop.

13.80 ± 1.53 12.62 ± 1.20 13.91 ± 1.09

11.82 ± 1.76 3.57 ± 1.38* 4.06 ± 1.66*

25.25 ± 6.69 60.22 ± 14.65t 112.71 ± 25.18t

19.57 ± 6.69 17.40 ± 5.76 24.36 ± 9.10

Legend: SEP, Somatosensory evoked potentials. AXC, Aortic cross-clamp. 'p < 0.0001 versus baseline. tp

< 0.02 versus

baseline.

Table VI. Group IIb: Changes in spinal cord blood flow (ml/l00 gmimin] in animals exhibiting 10 minutes of complete SEP loss (AXC time = 18 minutes) Spinal cord segment

Baseline

SEP loss

20 min after reperfusion

7 days postopd

c-r,

18.04 ± 1.94 15.35 ± 1.60 18.61 ± 2.27

12.62 ± 1.78 5.64 ± 1.96* 3.26 ± 1.16*

22.25 ± 3.50 46.53 ± 6.63* 69.01 ± 13.03t

9.94 ± 1.45 9.13 ± 1.95 13.93 ± 2.72

T6-13

Ll-s

'p < 0.001 versus baseline. tp < 0.005 versus baseline. tData shown are based on analysis (paired Student's t test) of five of nine animals (two spastic paraplegia, two normal, one ataxia: animals 3 and 7, 4 and 9, and 6, respectively, Table IV).

by a significant reactive hyperemic response characterized by a fourfold to sixfold increase in blood flow to these segments upon reperfusion. Physical separation between motor (ventral) and sensory (dorsal) cord pathways is a likely explanation for the observation that SEPs may return to normal

despite the presence of residual paraplegia.":" Preservation of small areas of the posterior columns of the spinal cord is sufficient to conduct normal SEP traces I9, 22.24 and appears to be the mechanism by which SEPs returned to normal at some point postoperatively in 60% of animals exhibiting permanent evidence of severe

Volume 94 Number 2 August 1987

neurologic injury in this study. Indeed, Coles and associates" have also shown that reduction in cord flow sufficient to produce isolated motor pathway infarction is associated with at least temporary ischemic sensory pathway dysfunction and SEP changes after aortic cross-clamping. In summary, brief periods of spinal cord ischemia (3 to 4 minutes) produce significant alterations in SEPs with rapid loss of SEPs if ischemia continues. Spinal cord ischemic intervals as brief as 18 minutes in the noncollateralized canine model are associated with a high incidence (67%) of spastic paraplegia. After reperfusion, infarction of cord tissue is best assessed by subsequent neurologic evaluation and may not always correlate with late measurements of spinal cord blood flow or the character ofSEP tracings. Early detection of intraoperative ischemic events by monitoring SEPs offers promise· in prevention of permanent spinal cord injury during surgical procedures on the thoracoabdominal aorta. REFERENCES I. Crawford ES, Fenstermacher JM, Richardson W, Sandiford F. Reappraisal of adjuncts to avoid ischemia in the treatment 'of thoracic aortic aneurysms. Surgery 1970;67:182-96. 2. Crawford ES, Waler HSJ III, Saleh SA, Normann NA. Graft replacement of aneurysm in descending thoracic aorta: results without bypass or shunting. Surgery 1981; 89:73-85. 3. Najafi H, Javid H, Hunter J, Serry C, Monson D. Descending aortic aneurysmectomy without adjuncts to avoid ischemia. Ann Thorac Surg 1980;30:326-35. 4. DeBakey ME, McCollum CH, Graham JM. Surgical treatment of aneurysms of the descending thoracic aorta: long-term results in 500 patients. J Cardiovasc Surg 1978;19:571-6. 5. Wolfe WG, Kleinman LH, Wechsler AS, Sabiston DC Jr. Heparin-coated shunts for lesions of the descending thoracic aorta. Arch Surg 1977;112:1481-7. 6. Lawrence GH, Hessel EA, Sauvage LR, Krause AH. Results of use of the TDMAC-heparin shunt in surgery of aneurysms of the descending thoracic aorta. J THORAC CARDIOVASC SURG 1977;73:393-8. 7. Donahoo JS, Brawley RK, Gott VL. The heparin-coated vascular shunt for thoracic aortic and great vessel procedures: a ten-year experience. Ann Thorac Surg 1977; 23:507-13. 8. May lA, Ecker RR, Iverson LIG. Heparinless femoral venoarterial bypass without an oxygenator for surgery on the descending thoracic aorta. J THORAC CARDIOVASC SURG 1977;73:387-92. 9. Wakabayashi A, Connolly JE. Prevention of paraplegia associated with resection of extensivethoracic aneurysms. Arch Surg 1976;111:1186-9.

Monitoring of SEPs. I

265

10. Wakabayashi A, Connolly JE. Heparinless left heart bypass for resection of thoracic aortic aneurysms. Am J Surg 1975;130:212-8. 11. Brewer LA, Fosburg RG, Mulder GA, Verska JJ. Spinal cord complications following surgery for coarctation of the aorta. J THoRAc CARDIOVASC SURG 1966;64:368-81. 12. Crawford ES, Snyder DM, Cho GC, Roehm JO Jr. Progress in the treatment of thoracoabdominal and abdominal aortic aneurysms involving celiac, superior mesenteric, and renal arteries. Ann Surg 1978;188:40422. 13. Adams HD, Van Geertruyden HH. Neurologic complications of aortic surgery. Ann Surg 1956;144:574-610. 14. Tschetter TH, Klassen AC, Resch JA, Meyer MW. Bloodflow in the central and peripheral nervous system in dogs using a particle distribution method. Stroke 1970; 1:370-4. 15. Laschinger JC, Cunningham IN Jr, Catinella FC, Nathan 1M, Knopp EA, Spencer Fe. Detection and prevention of intraoperative spinal cord ischemia after cross-clampingof the thoracic aorta: use of somatosensory evoked potentials. Surgery 1982;92: 1109-17. 16. Coles JG, Wilson GJ, Sima AF, Klement P, Tait GA. Intraoperative detection of spinal cord ischemia using somatosensoryevoked potentials. Ann Thorac Surg 1982; 34:299-308. 17. Tarlov 1M. Spinal cord compression: mechanism of paralysis and treatment. Springfield: Charles C Thomas, 1957:147. 18. Djindjian R, Hurth M, Houdart M, Laborit G, Julian H, Mamo H. Arterial supply of the spinal cord. In: Djindjian R, ed. Angiography of the spinal cord. Baltimore: Baltimore University Park Press, 1970:2-13. 19. Speilholz NI, Benjamin MV, Enger G, Ransohoff J. Somatosensory evoked potentials and clinical outcome in spinal cord injury. In: Popp AJ, ed. Neural trauma. New York: Raven Press, 1979:217-22. 20. Nash CL, Schatzinger L, Brown RH, Brodkey J. The unstable thoracic compression fracture: its problems and the use of spinal cord monitoring in the evaluation of treatment. Spine 1977;2:261-5. 21. Donaghy RM, Numoto M. Prognostic significance of somatosensory evoked potentials in spinal cord injury. Presented at the Seventeenth Spinal Cord Injury Conference. Veterans Administration Hospital, Bronx, New York, 1969. 22. D'Angelo CM, VanGilder JC, Taub A. Evoked cortical potentials in experimental spinal cord trauma. J Neurosurg 1973;38:332-6. 23. Cohen AR, Young W, Ransohoff J. Intraspinal Iocalization of SEP. Neurosurgery 1981;9:157-62. 24. Fried LC, Aparicio O. Experimental ischemia of the spinal cord: histologic studies after anterior spinal artery occlusion. Neurology 1973;23:289-93.