Local spinal cord blood flow in experimental traumatic myelopathy ARTHUR I. KOBRINE, M.D., THOMAS F. DOYLE, B.S., AND ALBERT N. MARTINS, M.D.
Neurosurgery Section, Walter Reed General Hospital, Washington, D.C., and Neurobiology Division, Armed Forces Radiobiology Research Institute, Bethesda, Maryland v' Focal blood flow was measured in the lateral funiculus and center of the spinal cord in the rhesus monkey both before and after a 600 gm-cm injury at T-10. Measurements made by the hydrogen clearance technique showed that blood flow in the lateral funiculus more than doubled within 4 hours after injury, returned to normal by 8 hours, and remained in the normal range for 24 hours. At no time was a hypoperfusion in the lateral funiculus present. Blood flow in the center of the spinal cord, at the level of the lesion, began to fall within 1 hour following injury and continued to fall for 4 hours. These data challenge the notion that spreading ischemia of the white matter is an important factor in the pathophysiology of experimental spinal cord injury. KEy WoRos
9 blood flow
T
HE prevailing theory concerning the pathophysiology of experimental spinal cord injury entails progressive ischemia in the lateral white matter, secondary to the development of a hemorrhagic infarct in the central gray matter, and possibly related to catecholamine release at the time of injury. 14 We have previously used the hydrogen clearance method to measure normal spinal cord blood flow (SCBF); 1~ in the present report, we use this method to measure focal blood flow changes in experimental spinal cord injury, in an attempt to test this theory. Materials and Methods Twenty adult rhesus monkeys (3 to 4 kg), unselected as to sex, were used in this study and divided into the following four groups.
]44
9 spinal cord injury
Group A The four animals in Group A were anesthetized with 0.5 ml phencyclidine H C I * and 1.0 ml sodium pentobarbital. A standard dorsal laminectomy was performed, and a 16gm weight was dropped 37.5 cm onto an aluminum anvil resting on the intact dura over the spinal cord at the T-10 level; this blow imparted 600 gm-cm to the cord. The footplate was contoured to saddle the spinal cord and had an area of 15.7 mm 2. The wound was closed, and these animals acted as chronic controls to determine the clinical extent of the neurological deficit with this lesion. *Phencyclidine HCI, Sernylan, manufactured by Biocentric Labs, Inc., St. Joseph, Missouri.
J. Neurosurg. / Volume 42 / February, 1975
S p i n a l c o r d b l o o d flow in t r a u m a t i c
myelopathy
Group B In these seven monkeys, multiple measurements of normal SCBF in the lateral funiculus were carried out during a 4-hour period under normotensive, normocapnic conditions. These animals were initially anesthetized with 0.5 ml phencyclidine HC1 and 1.0 ml sodium pentobarbital. Catheters were inserted into the femoral artery for continuous blood pressure monitoring and periodic blood gas determination, and into the femoral vein for fluid replacement at a rate of 16 ml/hour. The animals were then intubated, curarized, and placed on a respirator. They received N20 and 02 in a 2:1 mixture for the remainder of the experiment. Temperature, monitored by a rectal probe, was kept constant at 37 ~ to 39~ using a heating pad when necessary. The animals were hyperventilated to decrease the recirculation of hydrogen, and CO~ was added to the inspired air to maintain the arterial pCO~ constant at physiological levels of 30 to 35 mm Hg. 15 A dorsal laminectomy was performed exposing the dura mater over T7-11. One to three platinum electrodes, 250 u in diameter, were inserted 1 cm apart into the spinal cord through the intact dura at a point midway between the midline and lateral border, perpendicular to the surface; they were then advanced to a depth of 2 mm. This consistently placed the electrode tip in the lateral funiculus of the spinal cord. The hydrogen clearance apparatus used is similar to that reported by Aukland, et al., 1 but has been slightly modified in our laboratory to stabilize the baseline. The details of the circuitry have been described elsewhere. 1~ This technique measures blood flow in a discrete volume of tissue, less than 0.5 ramS.12 Blood flow measurements were carried out as follows. As 10% hydrogen gas was added to the inspired air for several minutes, the "wash in" of hydrogen into the spinal cord was monitored on the polygraph. Upon cessation of the hydrogen inhalation, the washout of hydrogen was recorded on the polygraph as a monoexponential decay curve, subsequently analyzed graphically by hand. It was also analyzed by a computer programmed for a linear regression analysis that gave the least squares best fit of the monoexponential equation to the washout curve. The flow was then calculated from the slope of the curve.
Group C The five animals in Group C were initially prepared as in Group B, and control flows were monitored for 1 hour. Then the middle electrode was removed and an injury as described above was delivered to the intact dura. The electrode was replaced in the lateral funiculus and blood flow was determined every half hour to every hour for 4 to 24 hours. At the end of this period, the animals were sacrificed, and the spinal cord removed for study.
J, Neurosurg. / Volume 42 / February, 1975
Group D The four animals in Group D were also prepared as in Group B; however, the platinum electrodes were inserted into the center of the spinal cord. After 1 hour of monitoring control values, as in Group C, the middle electrode was removed for trauma and then replaced. Flows were then obtained from all electrodes for 4 hours following trauma. At the end of this period, the animals were sacrificed, and the spinal cord rer~aoved for study. Results Systemic arterial blood pressure in all groups remained constant and in the physiological range throughout the entire experiment, except immediately after the spinal cord injury. Within 30 seconds following trauma, there was a rise of 25 to 30 mm Hg in all injured animals, which lasted 60 to 90 seconds before returning to the normal range. Figure 1 is a composite diagram of the experimental design, including the results.
Group A All of the animals in Group A were immediately rendered paraplegic, and remained so. Two animals lived for 7 days following injury and two for 14 days. Based on this observation and the reports of others, s we feel this injury always causes permanent paraplegia. Group B Based on 68 separate determinations from 12 electrodes in seven animals, the overall mean as measured in the lateral funiculus at T8-10 was 17.5 ml/min/100 gm + 0.346 sE (standard error). Four of these animals were allowed to awaken after the experimental period to undergo neurological examination; this was normal in all animals tested? ~ 145
A. I. K o b r i n e , T. F. D o y l e a n d A. N . M a r t i n s GROUP A
PERMANENTLY PARAPLEGIC
6 0 0 g m crn
INJURY TO TIO
GROUP B
NORMAL FLOW= 17.5ml/min/100gm~ + 0.346 S EM
NORMOTENSIVE NORMOCAPNIC GROUP
C
FLOW 130
ml/mm/lOOgtm 2o 10
6 0 0 g m cm INJURY
0
4.
600gm crn INJURY t ~ . ~ 1 ~ /
6
I 24
112
TIME(hours)
ml/mln/lOg~m 105 0
2 TIME (hours)
4
FIG. 1. Diagrammatic representation of the experimental design and results.
42 38 oio
34
EXPERIMENTAL A N I M A L S
....... 9 9
E
ANIMALS
O1
o ~ 30 e-
E 26
022 ,..i
u.
18
14 lO -lY2-~/20 1 '~ TRAUMA
2
3
4
5
6
8 10 12 14 TIME (hours) POST T R A U M A
16
18
20
22
24
Fl~. 2. Blood flow in the lateral funiculus at the level of trauma after a 600 gm-cm injury to T-10.
Group C In the animals from Group C, blood flow in the lateral funiculus at the level of trauma rose significantly within 1 hour following trauma, more than doubling the normal level (p < 0.01). Blood flow returned to the normal 146
range by 8 hours and remained in the normal range for 24 hours. At no time did the blood flow in the white matter at the level of trauma fall below the normal range (Fig. 2). Blood flow measured in the lateral funiculus 1 cm rostral and caudal to the level of injury followed a similar pattern, but to a somewhat J. Neurosurg. / Volume 42 / February, 1975
S p i n a l c o r d b l o o d f l o w in t r a u m a t i c 3Oil
o o =
O
II
I
I
I
I
I
I
I
I
[
I
I
I
I
I
26 22
14 LI"q 18 - ~T ~ / / i 10
I~'
2 3 4
5 6
8 10 12 14 TIME (hours) POST TRAUMA
16
18
20
22
24
10 -1]/2-1/20 1 2 3 4
5 6
8 10 12 14 TIME (h~t~,rs) POST TRAUMA
16
18
20
22
24
-172-'120
1
TRAUMA 30 ,r-v---v-~"~o o
I
myelopathy
261
= 22 E 18 O ._1 14
TRAUMA
Fie. 3. Blood flow in the lateral funiculus after a 600 gm-cm injury to T-10. Upper."Blood flow 1 cm rostral to the level of trauma. Lower."Blood flow 1 cm caudal to the level of trauma.
Discussion lesser degree (Fig. 3). Examination of the pathological specimen demonstrated a Our results indicate a need to reassess the hemorrhagic lesion in the center of the spinal popular pathophysiological theory stressing cord at the level of injury, which entirely the importance of lateral white matter replaced the gray matter. The spinal cord 1 ischemia in experimental spinal cord injury. cm rostral and caudal to the trauma site was Actually much in the previously reported exgrossly normal. perimental phenomena supports this reassessment. In all experimental designs, animals inGroup D jured with a force sufficient to render them During the control period in these four paraplegic have become so immediately, at a monkeys normal blood flo~) in the center of time when there has been minimal change in the cord was 14 ml/min/100 gm + 0.5 sE the light and electron microscopic examina(N = 27). Blood flow in the center of the cord tion of both the central gray and lateral white at the level of trauma fell within 1 hour after matter. 3,16 Moreover, the observed absence of injury, and continued to fall during the 4-hour sensory-evoked potentials ~ indicates that the experimental period (Fig. 4). Blood flow in spinal cord is totally unable to conduct imthe central cord 1 cm rostral and caudal to the pulses at this time. Studies correlating the injury site remained essentially unchanged. clinicopathological changes seen in varying Examination of the pathological specimen degrees of trauma have shown an inverse temdemonstrated a hemorrhagic lesion in the poral relationship between function and center of the spinal cord at the level of pathology; animals injured with a force causinjury, somewhat less extensive than that in ing transient paraplegia demonstrated a worsening of the central lesion as they were Group C. J. Neurosurg. / Volume 42 / February, 1975
147
A. I. Kobrine, T. F. Doyle and A. N. Martins FLOW ml/min/100 gm
18 16 L 14
10
1
4 2
0
I
1
I
2
I
3
4
I
5
TIME (hours) POST TRAUMA
Fl~. 4. Blood flow in the center of the spinal cord at the level of trauma after a 600 gm-cm injury to T-10.
clinically improving. 4 Hedeman and Sil, 9 unable to demonstrate any rise in norepinephrine levels of injured spinal cord, also found no correlation between the size or evolution of the central lesion and the return of neurological function in treated animals? Therefore, evidence exists that the well known, easily observed evolution of the central lesion may in effect be an epiphenomenon, caused either by the initial trauma or release of some as yet unknown substance, but with no cause and effect relationship to the devastating neurological deficit. Evidently, total dysfunction of the center of the cord at thoracic levels causes essentially no deficit in neurological function, since the ascending and descending tracts are situated in the peripheral white matter. Locke, et al.? 1 have demonstrated an increasing lactate level in the injured segment of spinal cord and considered this to be further evidence of spreading ischemia. However, since the entire segment was utilized in the lactate determination the elevated level could represent changes in the metabolism of cells in the central lesion rather than metabolic changes in the peripheral white matter. Our data from Group D are consistent with a significant ischemia in the central lesion following severe experimental spinal cord in148
jury, and would explain the elevated lactate levels. Furthermore, previous attempts to measure blood flow in the spinal cord by direct injection of xenon into the parenchyma of the cord e or by the washout of argon, measured by a 2-mm probe placed in the cord, 5 have encountered difficulties. These methods either create a lesion in the tissue in which flow is measured, or they are ill-suited to the measurement of focal flow, since they show no distinction between flow in the lateral white matter and flow in the central lesion. The marked rise in SCBF measured in the lateral funiculus after traumatic paraplegia in our experiment could result from hyperperfusion in response to an increased metabolic demand, or more probably from a luxury perfusion state, with loss of autoregulation and subsequent vascular dilatation as a direct result of trauma, or in relation to the nearby necrotic tissue in the center of the cord. Naftchi, et al., 18 demonstrated increased histamine levels at the site of experimental spinal cord injury as well as 1 cm rostral and caudal. This substance, which has been shown to cause vascular dilatation when injected,7 could be the causative agent to explain increased flow in the lateral funiculus. We would suggest an alternative explanation for the permanent neurological dysfunction seen in severe experimental spinal cord trauma. The hyperperfusion demonstrated in the lateral funiculus suggests that ischemia does not exist in this area following severe spinal cord trauma, and therefore plays no role in the associated pathophysiology. We suggest that the initial trauma immediately and permanently disrupts function of the axonal cell membrane in the lateral white matter, either by interfering with the selective permeability of the membrane to sodium and potassium or in some other way altering the ability of the axonal membrane to conduct an action potential. We further suggest that the lesion that evolves in the center of the injured segment, whether caused by direct trauma or the release of some unknown substance, is in effect an epiphenomenon and probably plays no role in the pathogenesis of the lateral white matter dysfunction. The return of function seen in experimental transient paraplegia is probably based on the return of functional integrity of the cell membrane rather than a change in blood flow. J. Neurosurg. / Volume 42 / February, 1975
Spinal cord blood flow in traumatic myelopathy Summary and Conclusions Focal blood flow was measured in the thoracic spinal cord both before and after trauma sufficient to render the animal permanently paraplegic. Blood flow measured in the center of the spinal cord at the level of the lesion fell within 1 hour following injury and continued to fall during the 4-hour period of measurement, while blood flow in the surrounding white matter increased 100% or more above the normal value (17.5 ml/min/100 gm) within several hours after injury, returned to the normal range by 8 hours, and remained in the normal range for 24 hours. No evidence from this experiment supports the view that trauma to the spinal cord is followed by a catecholamine-mediated spreading ischemia of the white matter. We have concluded, therefore, that traumatic paraplegia probably results from the functional disruption of axons at the instant of injury. We suggest that future experiments involving changes in metabolism or blood flow following spinal cord injury should employ only those techniques that have the capability of selectively measuring the variable in either the central lesion or the peripheral white matter since we now believe that two totally different mechanisms are probably operative.
6. 7. 8. 9.
10.
11. 12.
13.
14.
Acknowledgment The authors wish to express their appreciation to HM1 Walter Stringfield for his technical assistance.
15.
References 1. Aukland K, Bower BF, Berliner RW: Measurement of local blood flow with hydrogen gas. Circ Res 14:164-187, 1964 2. Campbell JL, DeCrescito V, Tomasula J J, et al: Bioelectric prediction of permanent posttraumatic paraplegia. Presented at the 39th Annual Meeting of the American Association of Neurological Surgeons, Houston, April, 1971 3. Dohrman G J, Wagner FC, Bucy PC: Transitory traumatic paraplegia: electron microscopy of early alterations in myelinated nerve fibers. J Neurosurg 36:407-415, 1972 4. Ducker TB, Kindt GW, Kempe LG: Pathological findings in acute experimental spinal cord trauma. J Neurosurg 35:700-708, 1971 5. Ducker TB, Lucas JT: Recovery from spinal cord injury, in Seeman P, Brown GM (eds): J. Neurosurg. / Volume 42 / February, 1975
16.
17.
Frontiers in Neurology. Toronto, University of Toronto Press, 1974, pp 142-154 Ducker TB, Perot PL Jr: Spinal cord oxygen and blood flow in trauma. Surg Forum 22: 413-415, 1971 Goodman LS, Gilman A: The Pharmacological Basis of Therapeutics. London, The Macmillan Co, 1969, p 616 Green B: Personal communication, 1974 Hedeman LS, Sil R: Studies in experimental spinal cord trauma. Part 2: Comparison of treatment with steroids, low molecular weight dextran, and catecholamine blockage. J Neurosurg 40:44-51, 1974 Kobrine AI, Doyle TF, Martins AN: Spinal cord blood flow in the Rhesus monkey by the hydrogen clearance method. Surg Neuroi 2:197-200, 1974 Locke GE, Yashon D, Feldman RA, et al: Ischemia in primate spinal cord injury. J Neurosurg 34:614-617, 1971 Meyer JS, Fukuuchi Y, Kanda T, et al: Regional measurements of cerebral.blood flow and metabolism using intracarotid injection of hydrogen, in Ross Russell RW (ed): International Symposium on the Regulation of Cerebral Blood Flow, 4th, London. London, Pitman Medical and Scientific Publishers, 1971, pp 71-80 Naftchi NE, Demeny M, DeCrescito V, et al: Biogenic amine concentrations in traumatized spinal cords of cats. Effect of drug therapy. J Neurosurg 40:52-57, 1974 Osterholm JL: The pathophysiological response to spinal cord injury. The current status of related research. J Neurosurg 40:5-33, 1974 Popovic NA, Mullane JF, Vick JA, et al: Effect of phencyclidine hydrochloride on certain cardiorespiratory values of the Rhesus monkey. (Macaca mulatta) Am J Vet Res 33:1649-1657, 1972 Wagner FC, Dohrman GJ, Bucy PC: Histopathology of transitory traumatic paraplegia in the monkey. J Neurosurg 35:272-276, 1971 Willis JA, Doyle TF, Ramirez A, et al: A Practical Circuit for Hydrogen Clearance Blood-Flow Measurement, bulletin # TN 74-2. Bethesda, Md, Armed Forces Radiobiology Research Institute Scientific Report, DNA, 1974
This paper was presented in part at the annual meeting of the American Association of Neurological Surgeons, April 24, 1974, St. Louis, Missouri. Address reprint requests to: Arthur I. Kobrine, M.D., Neurosurgery Section, Walter Reed General Hospital, Washington, D.C. 149
Neurosurgery Section, Walter Reed General Hospital, Washington, D.C., and Neurobiology Division, Armed Forces Radiobiology Research Institute, Bethesda, Maryland v' Focal blood flow was measured in the lateral funiculus and center of the spinal cord in the rhesus monkey both before and after a 600 gm-cm injury at T-10. Measurements made by the hydrogen clearance technique showed that blood flow in the lateral funiculus more than doubled within 4 hours after injury, returned to normal by 8 hours, and remained in the normal range for 24 hours. At no time was a hypoperfusion in the lateral funiculus present. Blood flow in the center of the spinal cord, at the level of the lesion, began to fall within 1 hour following injury and continued to fall for 4 hours. These data challenge the notion that spreading ischemia of the white matter is an important factor in the pathophysiology of experimental spinal cord injury. KEy WoRos
9 blood flow
T
HE prevailing theory concerning the pathophysiology of experimental spinal cord injury entails progressive ischemia in the lateral white matter, secondary to the development of a hemorrhagic infarct in the central gray matter, and possibly related to catecholamine release at the time of injury. 14 We have previously used the hydrogen clearance method to measure normal spinal cord blood flow (SCBF); 1~ in the present report, we use this method to measure focal blood flow changes in experimental spinal cord injury, in an attempt to test this theory. Materials and Methods Twenty adult rhesus monkeys (3 to 4 kg), unselected as to sex, were used in this study and divided into the following four groups.
]44
9 spinal cord injury
Group A The four animals in Group A were anesthetized with 0.5 ml phencyclidine H C I * and 1.0 ml sodium pentobarbital. A standard dorsal laminectomy was performed, and a 16gm weight was dropped 37.5 cm onto an aluminum anvil resting on the intact dura over the spinal cord at the T-10 level; this blow imparted 600 gm-cm to the cord. The footplate was contoured to saddle the spinal cord and had an area of 15.7 mm 2. The wound was closed, and these animals acted as chronic controls to determine the clinical extent of the neurological deficit with this lesion. *Phencyclidine HCI, Sernylan, manufactured by Biocentric Labs, Inc., St. Joseph, Missouri.
J. Neurosurg. / Volume 42 / February, 1975
S p i n a l c o r d b l o o d flow in t r a u m a t i c
myelopathy
Group B In these seven monkeys, multiple measurements of normal SCBF in the lateral funiculus were carried out during a 4-hour period under normotensive, normocapnic conditions. These animals were initially anesthetized with 0.5 ml phencyclidine HC1 and 1.0 ml sodium pentobarbital. Catheters were inserted into the femoral artery for continuous blood pressure monitoring and periodic blood gas determination, and into the femoral vein for fluid replacement at a rate of 16 ml/hour. The animals were then intubated, curarized, and placed on a respirator. They received N20 and 02 in a 2:1 mixture for the remainder of the experiment. Temperature, monitored by a rectal probe, was kept constant at 37 ~ to 39~ using a heating pad when necessary. The animals were hyperventilated to decrease the recirculation of hydrogen, and CO~ was added to the inspired air to maintain the arterial pCO~ constant at physiological levels of 30 to 35 mm Hg. 15 A dorsal laminectomy was performed exposing the dura mater over T7-11. One to three platinum electrodes, 250 u in diameter, were inserted 1 cm apart into the spinal cord through the intact dura at a point midway between the midline and lateral border, perpendicular to the surface; they were then advanced to a depth of 2 mm. This consistently placed the electrode tip in the lateral funiculus of the spinal cord. The hydrogen clearance apparatus used is similar to that reported by Aukland, et al., 1 but has been slightly modified in our laboratory to stabilize the baseline. The details of the circuitry have been described elsewhere. 1~ This technique measures blood flow in a discrete volume of tissue, less than 0.5 ramS.12 Blood flow measurements were carried out as follows. As 10% hydrogen gas was added to the inspired air for several minutes, the "wash in" of hydrogen into the spinal cord was monitored on the polygraph. Upon cessation of the hydrogen inhalation, the washout of hydrogen was recorded on the polygraph as a monoexponential decay curve, subsequently analyzed graphically by hand. It was also analyzed by a computer programmed for a linear regression analysis that gave the least squares best fit of the monoexponential equation to the washout curve. The flow was then calculated from the slope of the curve.
Group C The five animals in Group C were initially prepared as in Group B, and control flows were monitored for 1 hour. Then the middle electrode was removed and an injury as described above was delivered to the intact dura. The electrode was replaced in the lateral funiculus and blood flow was determined every half hour to every hour for 4 to 24 hours. At the end of this period, the animals were sacrificed, and the spinal cord removed for study.
J, Neurosurg. / Volume 42 / February, 1975
Group D The four animals in Group D were also prepared as in Group B; however, the platinum electrodes were inserted into the center of the spinal cord. After 1 hour of monitoring control values, as in Group C, the middle electrode was removed for trauma and then replaced. Flows were then obtained from all electrodes for 4 hours following trauma. At the end of this period, the animals were sacrificed, and the spinal cord rer~aoved for study. Results Systemic arterial blood pressure in all groups remained constant and in the physiological range throughout the entire experiment, except immediately after the spinal cord injury. Within 30 seconds following trauma, there was a rise of 25 to 30 mm Hg in all injured animals, which lasted 60 to 90 seconds before returning to the normal range. Figure 1 is a composite diagram of the experimental design, including the results.
Group A All of the animals in Group A were immediately rendered paraplegic, and remained so. Two animals lived for 7 days following injury and two for 14 days. Based on this observation and the reports of others, s we feel this injury always causes permanent paraplegia. Group B Based on 68 separate determinations from 12 electrodes in seven animals, the overall mean as measured in the lateral funiculus at T8-10 was 17.5 ml/min/100 gm + 0.346 sE (standard error). Four of these animals were allowed to awaken after the experimental period to undergo neurological examination; this was normal in all animals tested? ~ 145
A. I. K o b r i n e , T. F. D o y l e a n d A. N . M a r t i n s GROUP A
PERMANENTLY PARAPLEGIC
6 0 0 g m crn
INJURY TO TIO
GROUP B
NORMAL FLOW= 17.5ml/min/100gm~ + 0.346 S EM
NORMOTENSIVE NORMOCAPNIC GROUP
C
FLOW 130
ml/mm/lOOgtm 2o 10
6 0 0 g m cm INJURY
0
4.
600gm crn INJURY t ~ . ~ 1 ~ /
6
I 24
112
TIME(hours)
ml/mln/lOg~m 105 0
2 TIME (hours)
4
FIG. 1. Diagrammatic representation of the experimental design and results.
42 38 oio
34
EXPERIMENTAL A N I M A L S
....... 9 9
E
ANIMALS
O1
o ~ 30 e-
E 26
022 ,..i
u.
18
14 lO -lY2-~/20 1 '~ TRAUMA
2
3
4
5
6
8 10 12 14 TIME (hours) POST T R A U M A
16
18
20
22
24
Fl~. 2. Blood flow in the lateral funiculus at the level of trauma after a 600 gm-cm injury to T-10.
Group C In the animals from Group C, blood flow in the lateral funiculus at the level of trauma rose significantly within 1 hour following trauma, more than doubling the normal level (p < 0.01). Blood flow returned to the normal 146
range by 8 hours and remained in the normal range for 24 hours. At no time did the blood flow in the white matter at the level of trauma fall below the normal range (Fig. 2). Blood flow measured in the lateral funiculus 1 cm rostral and caudal to the level of injury followed a similar pattern, but to a somewhat J. Neurosurg. / Volume 42 / February, 1975
S p i n a l c o r d b l o o d f l o w in t r a u m a t i c 3Oil
o o =
O
II
I
I
I
I
I
I
I
I
[
I
I
I
I
I
26 22
14 LI"q 18 - ~T ~ / / i 10
I~'
2 3 4
5 6
8 10 12 14 TIME (hours) POST TRAUMA
16
18
20
22
24
10 -1]/2-1/20 1 2 3 4
5 6
8 10 12 14 TIME (h~t~,rs) POST TRAUMA
16
18
20
22
24
-172-'120
1
TRAUMA 30 ,r-v---v-~"~o o
I
myelopathy
261
= 22 E 18 O ._1 14
TRAUMA
Fie. 3. Blood flow in the lateral funiculus after a 600 gm-cm injury to T-10. Upper."Blood flow 1 cm rostral to the level of trauma. Lower."Blood flow 1 cm caudal to the level of trauma.
Discussion lesser degree (Fig. 3). Examination of the pathological specimen demonstrated a Our results indicate a need to reassess the hemorrhagic lesion in the center of the spinal popular pathophysiological theory stressing cord at the level of injury, which entirely the importance of lateral white matter replaced the gray matter. The spinal cord 1 ischemia in experimental spinal cord injury. cm rostral and caudal to the trauma site was Actually much in the previously reported exgrossly normal. perimental phenomena supports this reassessment. In all experimental designs, animals inGroup D jured with a force sufficient to render them During the control period in these four paraplegic have become so immediately, at a monkeys normal blood flo~) in the center of time when there has been minimal change in the cord was 14 ml/min/100 gm + 0.5 sE the light and electron microscopic examina(N = 27). Blood flow in the center of the cord tion of both the central gray and lateral white at the level of trauma fell within 1 hour after matter. 3,16 Moreover, the observed absence of injury, and continued to fall during the 4-hour sensory-evoked potentials ~ indicates that the experimental period (Fig. 4). Blood flow in spinal cord is totally unable to conduct imthe central cord 1 cm rostral and caudal to the pulses at this time. Studies correlating the injury site remained essentially unchanged. clinicopathological changes seen in varying Examination of the pathological specimen degrees of trauma have shown an inverse temdemonstrated a hemorrhagic lesion in the poral relationship between function and center of the spinal cord at the level of pathology; animals injured with a force causinjury, somewhat less extensive than that in ing transient paraplegia demonstrated a worsening of the central lesion as they were Group C. J. Neurosurg. / Volume 42 / February, 1975
147
A. I. Kobrine, T. F. Doyle and A. N. Martins FLOW ml/min/100 gm
18 16 L 14
10
1
4 2
0
I
1
I
2
I
3
4
I
5
TIME (hours) POST TRAUMA
Fl~. 4. Blood flow in the center of the spinal cord at the level of trauma after a 600 gm-cm injury to T-10.
clinically improving. 4 Hedeman and Sil, 9 unable to demonstrate any rise in norepinephrine levels of injured spinal cord, also found no correlation between the size or evolution of the central lesion and the return of neurological function in treated animals? Therefore, evidence exists that the well known, easily observed evolution of the central lesion may in effect be an epiphenomenon, caused either by the initial trauma or release of some as yet unknown substance, but with no cause and effect relationship to the devastating neurological deficit. Evidently, total dysfunction of the center of the cord at thoracic levels causes essentially no deficit in neurological function, since the ascending and descending tracts are situated in the peripheral white matter. Locke, et al.? 1 have demonstrated an increasing lactate level in the injured segment of spinal cord and considered this to be further evidence of spreading ischemia. However, since the entire segment was utilized in the lactate determination the elevated level could represent changes in the metabolism of cells in the central lesion rather than metabolic changes in the peripheral white matter. Our data from Group D are consistent with a significant ischemia in the central lesion following severe experimental spinal cord in148
jury, and would explain the elevated lactate levels. Furthermore, previous attempts to measure blood flow in the spinal cord by direct injection of xenon into the parenchyma of the cord e or by the washout of argon, measured by a 2-mm probe placed in the cord, 5 have encountered difficulties. These methods either create a lesion in the tissue in which flow is measured, or they are ill-suited to the measurement of focal flow, since they show no distinction between flow in the lateral white matter and flow in the central lesion. The marked rise in SCBF measured in the lateral funiculus after traumatic paraplegia in our experiment could result from hyperperfusion in response to an increased metabolic demand, or more probably from a luxury perfusion state, with loss of autoregulation and subsequent vascular dilatation as a direct result of trauma, or in relation to the nearby necrotic tissue in the center of the cord. Naftchi, et al., 18 demonstrated increased histamine levels at the site of experimental spinal cord injury as well as 1 cm rostral and caudal. This substance, which has been shown to cause vascular dilatation when injected,7 could be the causative agent to explain increased flow in the lateral funiculus. We would suggest an alternative explanation for the permanent neurological dysfunction seen in severe experimental spinal cord trauma. The hyperperfusion demonstrated in the lateral funiculus suggests that ischemia does not exist in this area following severe spinal cord trauma, and therefore plays no role in the associated pathophysiology. We suggest that the initial trauma immediately and permanently disrupts function of the axonal cell membrane in the lateral white matter, either by interfering with the selective permeability of the membrane to sodium and potassium or in some other way altering the ability of the axonal membrane to conduct an action potential. We further suggest that the lesion that evolves in the center of the injured segment, whether caused by direct trauma or the release of some unknown substance, is in effect an epiphenomenon and probably plays no role in the pathogenesis of the lateral white matter dysfunction. The return of function seen in experimental transient paraplegia is probably based on the return of functional integrity of the cell membrane rather than a change in blood flow. J. Neurosurg. / Volume 42 / February, 1975
Spinal cord blood flow in traumatic myelopathy Summary and Conclusions Focal blood flow was measured in the thoracic spinal cord both before and after trauma sufficient to render the animal permanently paraplegic. Blood flow measured in the center of the spinal cord at the level of the lesion fell within 1 hour following injury and continued to fall during the 4-hour period of measurement, while blood flow in the surrounding white matter increased 100% or more above the normal value (17.5 ml/min/100 gm) within several hours after injury, returned to the normal range by 8 hours, and remained in the normal range for 24 hours. No evidence from this experiment supports the view that trauma to the spinal cord is followed by a catecholamine-mediated spreading ischemia of the white matter. We have concluded, therefore, that traumatic paraplegia probably results from the functional disruption of axons at the instant of injury. We suggest that future experiments involving changes in metabolism or blood flow following spinal cord injury should employ only those techniques that have the capability of selectively measuring the variable in either the central lesion or the peripheral white matter since we now believe that two totally different mechanisms are probably operative.
6. 7. 8. 9.
10.
11. 12.
13.
14.
Acknowledgment The authors wish to express their appreciation to HM1 Walter Stringfield for his technical assistance.
15.
References 1. Aukland K, Bower BF, Berliner RW: Measurement of local blood flow with hydrogen gas. Circ Res 14:164-187, 1964 2. Campbell JL, DeCrescito V, Tomasula J J, et al: Bioelectric prediction of permanent posttraumatic paraplegia. Presented at the 39th Annual Meeting of the American Association of Neurological Surgeons, Houston, April, 1971 3. Dohrman G J, Wagner FC, Bucy PC: Transitory traumatic paraplegia: electron microscopy of early alterations in myelinated nerve fibers. J Neurosurg 36:407-415, 1972 4. Ducker TB, Kindt GW, Kempe LG: Pathological findings in acute experimental spinal cord trauma. J Neurosurg 35:700-708, 1971 5. Ducker TB, Lucas JT: Recovery from spinal cord injury, in Seeman P, Brown GM (eds): J. Neurosurg. / Volume 42 / February, 1975
16.
17.
Frontiers in Neurology. Toronto, University of Toronto Press, 1974, pp 142-154 Ducker TB, Perot PL Jr: Spinal cord oxygen and blood flow in trauma. Surg Forum 22: 413-415, 1971 Goodman LS, Gilman A: The Pharmacological Basis of Therapeutics. London, The Macmillan Co, 1969, p 616 Green B: Personal communication, 1974 Hedeman LS, Sil R: Studies in experimental spinal cord trauma. Part 2: Comparison of treatment with steroids, low molecular weight dextran, and catecholamine blockage. J Neurosurg 40:44-51, 1974 Kobrine AI, Doyle TF, Martins AN: Spinal cord blood flow in the Rhesus monkey by the hydrogen clearance method. Surg Neuroi 2:197-200, 1974 Locke GE, Yashon D, Feldman RA, et al: Ischemia in primate spinal cord injury. J Neurosurg 34:614-617, 1971 Meyer JS, Fukuuchi Y, Kanda T, et al: Regional measurements of cerebral.blood flow and metabolism using intracarotid injection of hydrogen, in Ross Russell RW (ed): International Symposium on the Regulation of Cerebral Blood Flow, 4th, London. London, Pitman Medical and Scientific Publishers, 1971, pp 71-80 Naftchi NE, Demeny M, DeCrescito V, et al: Biogenic amine concentrations in traumatized spinal cords of cats. Effect of drug therapy. J Neurosurg 40:52-57, 1974 Osterholm JL: The pathophysiological response to spinal cord injury. The current status of related research. J Neurosurg 40:5-33, 1974 Popovic NA, Mullane JF, Vick JA, et al: Effect of phencyclidine hydrochloride on certain cardiorespiratory values of the Rhesus monkey. (Macaca mulatta) Am J Vet Res 33:1649-1657, 1972 Wagner FC, Dohrman GJ, Bucy PC: Histopathology of transitory traumatic paraplegia in the monkey. J Neurosurg 35:272-276, 1971 Willis JA, Doyle TF, Ramirez A, et al: A Practical Circuit for Hydrogen Clearance Blood-Flow Measurement, bulletin # TN 74-2. Bethesda, Md, Armed Forces Radiobiology Research Institute Scientific Report, DNA, 1974
This paper was presented in part at the annual meeting of the American Association of Neurological Surgeons, April 24, 1974, St. Louis, Missouri. Address reprint requests to: Arthur I. Kobrine, M.D., Neurosurgery Section, Walter Reed General Hospital, Washington, D.C. 149
