J Neurosurg 74:270-277, 1991
Evaluation of brain-stem dysfunction following severe fluid-percussion head injury to the cat KATSUJI SHIN4A, M.D., D.M.Sc., AND ANTHONY MARMAROU, PH.D. Division of Neurosurgery, Medical College of Virginia, Richmond, Virginia v The degree of brain-stem dysfunction associated with high-level fluid-percussion injury (3.0 to 3.8 atm) was investigated in anesthetized cats. Measurements were made of the animals' intracranial pressure (ICP), pressure-volume index (FYI), far-field brain-stem auditory evoked responses (BAER's), and cerebral blood flow (CBF). The animals were classified into two groups based on the severity of neuropathological damage to the brain stem after trauma: Group 1 had mild intraparenchymal and subarachnoid hemorrhages and Group 2 had severe intraparenchymal and subarachnoid hemorrhages. The ICP values in Group 1 were insignificantly lower than those in Group 2, while the PVI values in Group 2 were clearly lower (p < 0.05). Immediately after the injury, peaks H, III, and IV of the BAER's demonstrated a transitory and marked suppression. One Group 1 and two Group 2 animals showed the disappearance of peak V. In Group 1, the latencies of peak II, III and IV gradually increased until 60 to 150 minutes postinjury, then returned to 95% of baseline value at 8 hours; however, the animals in Group 2 showed poor recovery of latencies. Two hours after brain injury, the CBF decreased to 40% of the preinjury measurement in both groups (p < 0.001). In contrast to Group 2, the CBF in Group I returned to 86.8% of the preinjury measurement by 8 hours following the injury. Changes in PVI, BAER, and CBF correlated well with the degree of brain-stem injury following severe head injury. These data indicate that high-level fluid-percussion injury (> 3.0 atm) is predominantly a model of brain-stem injury. ,
KEY WORDS • head injury • pressure-volume index • fluid-percussion injury • brain-stem auditory evoked response • cerebral blood flow • cat
T
model has been used widely in the production of experimental head injury. Investigators using this model have made substantial contributions to the body of knowledge elucidating mechanisms of brain trauma. In recent studies, conducted at high injury levels, we observed that the edema resulting from severe experimental brain injury in animals that died shortly after injury was maximal in brain-stem areas.' We hypothesized that animals subjected to high levels of fluid percussion, which involves significant fluid volume injection into the closed calvaria, may be predisposed to severe brain-stem damage when tissue is displaced and the structural threshold is exceeded. This might explain the bimodel distribution in mortality seen frequently in animals subjected to high levels of fluid-percussion injury. The evidence from clinical" .12 . 35 and experimental' 26.34 neuropathology indicates that the degree of brainstem damage following severe head trauma is one of the most important factors influencing outcome. Adams, et al., 3 reported that the rostral aspect of the brain 270
HE fluid-percussion
stem is a common site for the most severe degree of diffuse axonal injury and of multiple petechial hemorrhaging, which is found frequently in patients who die very shortly after head trauma. Heretofore, the characteristics of the fluid-percussion model at high levels of injury have not been addressed. Knowledge of both the mechanical behavior and pathology produced at high injury levels is essential to the study of lethal percussion injury and the similarity of experimental and clinical observations. Thus, the first objective of the present study was to detect and quantitatively describe the severity of brain-stem damage following high-level fluid-percussion injury utilizing both neurophysiological analysis with the brain-stem auditory evoked response (BAER) and brain compliance with the pressure-volume index (PVI). These measurements have been shown to be sensitive indicators of brain-stern dysfunction and brain swelling. 5. 35 9,12-15 - 2 °. ." With the pathophysiological response to severe fluid percussion defined, the second objective was to study the structural response. This was accomplished J. Neurosurg_ / Volume 74 /February, 1991
Brain-stem dysfunction after fluid-percussion head i njury in collaboration with the bioengineering laboratories of the University of Pennsylvania and will be described in a separate report.
Materials and Methods Adult cats, weighing between 2.5 and 4.0 kg, were anesthetized using intravenous rnethohexital sodium titrated to the absence of blink reflex (2 to 5 ml of a 1% solution). Following initial induction of anesthesia, the animals were ventilated with a gas mixture containing 70% NO and 30% 0 2 coupled with a supplemental barbiturate (35 mg/kg) sufficient to maintain surgical depth of anesthesia for cannulation of the femoral artery and vein and cranial preparation for the injury coupling device. During these procedures, gallamine triethiodide (4 mg/kg) was used to maintain muscle relaxation for mechanical ventilation and a reliable endtidal CO 2 concentration of 4% to 4.5% (arterial pCO 2 between 26 and 32 mm Hg). During the entire experimental period, adequate depth of anesthesia was assessed by observing the blood pressure and heart rate response to paw pinch. A heating pad was used to maintain body temperature at 37° to 39°C. Following these procedures, the animals were placed on a stereotactic frame and maintained in a sphinx position with hollow ear bars, avoiding damage to the auditory canal that might interfere with accurate BAER measurement. After the scalp was reflected, a craniotomy 11 mm in diameter was made over the left temporal lobe just rostra] to the external auditory meatus, and a hollow metal screw was tightly inserted in the cranium over the intact dura. Bilateral small craniotomies, 3 mm in diameter, were made in the frontal bone prior to brain injury and in the parietal bone immediately following the injury for insertion of Teflon-coated platinum electrodes for the measurement of cerebral blood flow (CBF). A stainless-steel screw was secured at the vertex for measuring far-field BAER's, and a second reference electrode was inserted into the inner pinna of the right ear. Bifrontal screws for electroencephalography (EEG) were also affixed to the bone. An additional two electrodes at the brow and the occipital protuberance served as grounds for BAER measurement and EEG, respectively.
Measurement of BAER The BAER's were recorded by means of an averager.* Stimulation was by 100 Asec square-wave clicks delivered at 10.17sec and 75 dB sound pressure level intensity through the polyethylene (PE) tubing (12 cm) connecting the earphone to the external canal. The positive input was obtained from the vertex. Recording electrodes were connected to a preamplifier (150 to 3000 Hz bandpass, 104 gain). The amplified responses were * Averager, Model CA 1000, manufactured by Nicolet Instrument Corp., Madison, Wisconsin.
Neurosurg. / Volume 74 / February, 1991
averaged 500 times at the analysis time of 10 cosec, and charted on an x-y recorder.
Measurement of Local CBF Local CBF was measured with the hydrogen-clearance technique.' Teflon-coated platinum needles, 254 /./ in diameter, were implanted under the operating microscope into the gray matter of the right and left suprasylvian gyrus according to stereotactic coordinates (preinjury: +16 AP, 8.5 Lat; postinjury: —4 AP, 7.5 Lat). The reference was placed in the nuchal muscle. Hydrogen gas was administered into the inspired gas mixture and adjusted to the constant flow rate of 2%. The electrodes were connected to electrode current amplifiers and the output was then routed to a computer for on-line computation of CBF. The first measurement of CBF was not made until 1 hour after initial electrode placement in order to permit full polarization and stabilization of the electrodes.
Measurement of ICP and PVI Following the injury, the cisterna magna was cannulated with a No. 21 needle with an interposed fourway stopcock array which allowed continuous monitoring of intracranial pressure (ICP) and bolus injection for PVI. The needle was connected to a Statham pressure transducer (P23ID) through PE-50 tubing. Mock cerebrospinal fluid (0.1 ml) was injected into the cisterna magna for the determination of PVI to assess the changes in tissue compliance.'
Experimental Protocol After surgical preparation and before brain injury, the cats were removed from the stereotactic frames and baseline local CBF and BAER were measured. Brain injury was produced by a fluid-percussion device' attached securely to a right-angled hollow screw positioned in the temporal region to produce a lateral injury. High percussion levels between 3.0 and 3.8 atm were selected to produce severe brain injury, and the level of injury for each animal was measured from a pressure pulse recorded on an oscilloscope. Immediately following the injury, the injury tube was removed, and the craniotomy defect was first tilled with Gelfoam then sealed with dental cement. The BAER's were recorded at approximately 5-minute intervals until 30 minutes had passed, and then at 30-minute intervals until the end of each experiment. Local CBF was measured every 1 or 2 hours, while FYI was measured at 30-minute intervals beginning 2 hours after the brain injury. Thirty minutes before sacrifice by intravenous injection of KC1, each cat was injected with a 2% filtrated solution of Evans blue dye. At 15 minutes before sacrifice, the animals were injected with sodium pentobarbital (100 mg/kg) to anesthetize the animals deeply prior to ICI injection. Following sacrifice, macroscopical pathological analysis was carried out in all animals to assess the location and degree of brain injury and Evans blue staining. 271
K. Shima and A. Marmarou
FIG. l . The time course of mean arterial blood pressure (MABP, upper), pressure-volume index (PVI, center), and intracranial pressure (ICP, lower) in five cats with mild brainstem injury (Group 1, left) and in seven cats with severe brainstem injury (Group 2, right) following severe brain injury. Values are mean ± standard deviation. Type = animal group.
FIG. 2. Percent changes in amplitude of each peak of the brain-stem auditory evoked responses (BAER's) in Group 1 (lei) and Group 2 (right) animals following severe head injury. Roman numerals denote peaks in the brain-stem auditory evoked response.
Results
averaged 17.1 ± 3.64 mm Hg and increased mildly to an average of 19.8 ± 6.11 mm Hg at 8 hours. The ICP profiles in these two groups were not significantly different (Fig. 1). Although the ICP profiles among Group 1 and 2 animals were similar, the degree of brain swelling as indexed by PVI was significantly different. The PVI's in Group 1 averaged 0.978 ± 0.132 ml at 2 hours postinjury and 0.756 ± 0.033 ml at 8 hours (Fig. 1). However, in Group 2 animals, the initial PVI's measured at 2 hours were lower, averaging 0.615 ± 0.075 ml, then decreased to 0.498 ± 0.065 ml at 8 hours postinj ury.
The animals were classified into two groups based on the neuropathological changes in the brain stem. Group 1 (five cats) exhibited spot-like or small (< 3 aim in diameter) isolated intraparenchymal hemorrhages and subarachnoid hemorrhages, and Group 2 (seven cats) had diffuse or large (> 3 mm in diameter) multiple intraparenchymal hemorrhages and subarachnoid hemorrhages. The mean percussion injury levels (± standard deviation) were 3.2 ± 0.19 atm for Group 1 and 3.3 ± 0.34 atm for Group 2.
Physiological Changes In the 1st minute after injury, a loss of the corneal and pupillary reflex and a rapid increase in systemic mean arterial blood pressure (MABP) was observed. At 1 minute following the injury, the MABP was significantly increased by 97.0% in Group 1 and 81.6% in Group 2 over each baseline value (p < 0.001); however, there was no significant difference in peak MABP between Groups 1 and 2 (Fig. 1). After the initial rise, the MABP in Group I fell below baseline, then rapidly increased, reaching preinjury levels by 2 hours postimpact. Following these transient changes, the MABP remained at baseline throughout the 8 hours of observation. The initial transient changes of MABP in Group 2 animals were similar to those in Group 1 with the exception that, after crossing baseline, MABP continued to decrease gradually to baseline without recovery (Fig. 1 right). One animal in Group 2 was sacrificed 3.4 hours after injury because of a rapid decrease in MABP.
ICP and PT/1 Responses to Injury Cannulation of the cisterna magna was possible at 2 hours postinjury for measurement of ICP and PVI. In Group 1 animals, the initial ICP averaged 14.25 ± 2.86 mm Hg and remained at that level for the duration of the experiment. In Group 2 animals, the initial ICP 272
Brain-Stern Auditory Evoked Response The first five positive peaks of the BAER's were labeled sequentially from peak I to peak V. Immediately following brain injury, peaks II, III, and IV demonstrated a transient and marked suppression. The amplitude of peak I remained suppressed by 50% in each group except for one Group 2 cat at 2 hours after trauma (Fig. 2). In contrast, one Group 1 and all Group 2 cats showed the disappearance of peak V. In one Group 2 cat peaks III and IV also disappeared during the 8 hours after the injury; however, the amplitude of peak V in Group I animals recovered to 58.4% of the baseline value by 8 hours after trauma, which was close to the value of peak I at 8 hours. Peaks II, III, and IV in Group 1 cats remained at a milder reduction of amplitude than those in Group 2 throughout the experiment. All peaks in Group 1 animals showed a tendency to recover from the lowest values at 4 or 5 hours after the injury. In Group 2, however, each peak was markedly suppressed at a constant level below 40%. Individual examples of the effects of trauma upon the BAER's shown in Fig. 3 are characteristic of each group. The mean baseline latency (± standard deviation) of each peak and range was as follows: peak I, 1.86 ± 0.24 msec; peak II, 2.78 ± 0.17 msec; peak HI, 3.51 ± 0.17 J. Neurosurg. / Volume 74 / February, 1991
Brain-stem dysfunction after fluid-percussion head injury
FIG. 3. Examples of brain-stem auditory evoked responses (BAER's). Roman numerals denote BAER peaks. Left: Two examples of BAER's in Group 1 animals which recovered by 8 hours postinjury. The amplitude and latencies of all waves are not seriously affected. Right: Two examples of BAER's in Group 2 animals showing disappearance of early waves and specifically wave V.
msec; peak IV, 4.43 ± 0.18 msec; and peak V, 5.70 ± 0.39 msec (Fig. 4). In Group 1, the latency of peaks II, III, and IV was prolonged gradually until 60 to 150 minutes postinjury, then returned to approximately 95 % of the baseline value by 8 hours. In contrast, the animals in Group 2 did not show recovery of the latencies of all peaks.
Local Cerebral Blood Flow The preinjury local CBF values in the frontal cortex were 43.3 ± 3.3 m1/100 gm/min (4.2 ± 0.66 m1/100 gm/min in Group 1 cats and 42.7 ± 3.1 ml/gm/min in Group 2 cats). Two hours after the injury, the CBF significantly decreased to approximately 40% from the preinjury levels in each hemisphere in both groups (p < 0.001). The CBF in Group 1 animals gradually returned to 86.8% of the preinjury level by 8 hours
FIG. 4. Sequential changes in the latencies of the brainstem auditory evoked responses (BAER's) in five Group 1 (left) and seven Group 2 (right) animals following severe head injury. Roman numerals denote BAER peaks. Values are mean ± standard deviation. Wave V was absent following injury. Type = animal group.
J. Neurosurg. / Volume 74 / February, 1991
following trauma (Fig. 5). However, the low CBF in Group 2 cats persisted in each hemisphere until the end of the experiment.
Neuropathological Findings The only extravasation of Evans blue dye in the brains of these animals was a heterogeneous leakage in the cerebral cortex under the fluid-percussion tube. All of the cats showed moderate to severe subarachnoid hemorrhage. By classification, the five Group I cats had spot-like or small (< 3 mm in diameter) multiple intraparenchymal hemorrhages in the midbrain to the medulla oblongata. Six of the seven Group 2 animals had multiple petechial hemorrhages in the midbrain and pons, and even in the white matter of both cerebral
FIG. 5. Cerebral blood flow (CBF) in the left (injured) cerebral cortex and the right cerebral cortex following severe head injury. Circles indicate the mean CBF in four Group 1 animals; squares indicate the mean CBF in four Group 2 animals. Values are given ± standard deviation. * = p < 0.05; ** = p < 0.01; *** = p < 0.001; + = p < 0.05 vs. Group 1.
273
K. Shima and A. Marmarou
FIG. 7. Mean sequential changes in interpeak latencies (IPL) of brain-stem auditory evoked response peaks I to III (squares), III to V (circles), and I to V (triangles) in five Group 1 animals. Values are given ± standard deviation.
FIG. 6. Spatial distribution of brain-stem hemorrhages (> 1 mm) in five Group I (upper) and seven Group 2 (lower) animals following severe head injury.
hemispheres (Fig. 6). During the experimental period
there were no significant changes in p0 2 , pCO 2 , or pH. Discussion The results of our studies indicate that high-level fluid percussion is predominantly a brain-stem injury. The neurophysiological changes, specifically the absence of BAER wave V, points to structural damage of the midbrain and upper pons, which is in agreement with the morphological findings. Permanent damage, as assessed by failure of wave V to return, was correlated with sustained hypotension, low brain compliance and reduced CBF. With less tissue damage, both latency and amplitude of the BAER's recovered by 8 hours postinjury. Neurophysiological Correlates: Brain-Stem Injury
Many investigators have attempted to establish a relationship between the changes in the BAER's and brain-stem lesions confirmed by neurological and radiological findings. ' 2 . 14 . 35 ' 37 However, few attempts have been made to evaluate the sequential changes in brainstem dysfunction immediately following severe brain injury. In the current studies, although all BAER peaks demonstrated prolonged latencies and suppressed amplitudes following trauma, these alterations varied for each peak. Greenberg, et a1., 12 established a grading system of abnormal BAER's in patients with severe head trauma. Based on the degree of BAER abnormality, they noted delays of interpeak latency between peaks I and V in 17 of 21 patients. In their Grade III BAER, considered severely abnormal, only peak I has a normal latency and amplitude. Tsubokawa, et al.,' analyzed the BAER's in 64 274
patients within 3 days following severe head injury and divided the cases into two groups. One group had a prolonged latency or the disappearance of peak V, only six of the 13 patients in this group had a good outcome. In the present experimental study, the disappearance of peak V was observed in 10 of 12 cats within 3 minutes posttrauma. Wave V disappeared and did not recover in any cat with greater tissue damage as assessed by an increased number of petechial hemorrhages (Group 2). The absence of peak V suggests damage to the inferior colliculus which is thought to be the generating locus of peak By contrast, peak V in animals with less tissue damage (Group 1) appeared by approximately 20 minutes after injury and interpeak latencies of peaks I to V and peaks III to V, which are an index of central conduction time, returned to the baseline values by 8 hours (Fig. 7), It is of interest that the I to III interpeak latency did not show any change throughout the experiment. Anderson, et al.,' who viewed the I to V interpeak latency as a central auditory conduction time, reported that of 23 surviving patients with severe head trauma nine had the prolongation of I to V interpeak latencies. In our animals, the I to III interpeak latencies did not change during the prolongation of the I to V and II to V interpeak latencies. By contrast, in the patients with brain-stem stroke, the prolongation of the to III interpeak latencies was seen with normal III to V interpeak latencies in the patients with pontomedullary junction infarctions.' At high levels of injury, interpeak latency measures are of limited use since the BAER is obliterated. Thus, the use of latencies between peaks I and III and between peaks III and V is limited to evaluating brain-stem dysfunction and localizing lesions in the acute stage."' Survival With Experimental Brain Stem Injury
Following severe brain injury, the brain stem can be affected directly by a variety of hemorrhages, and later secondarily by uncal herniation. We have recently observed that cats, which were monitored for a 24-hour period following temporal and central fluid-percussion J. Neurosurg. / Volume 74 / February, 1991
Brain-stem dysfunction after fluid-percussion head injury trauma at the level of 2.8 atm, were classified into two groups based upon length of survival. Mean survival durations of the short-term survivors after central (impact on the vertex) and temporal (impact site over the posterior ectosylvian gyrus) impacts were 10.8 and 12.5 hours, respectively: rapid irreversible hypotension began after about 7 hours postinjury in both temporal and central injury.' Moreover, survival time was diminished in those animals with greater reduction in PVI. We suspect that brain-stem dysfunction played an important role in the outcome of these experiments, as the posttraumatic edema was maximal in brain-stern regions. Blood Pressure Alterations Transient hypertension followed by hypotension immediately after brain injury has been well documented in fluid-percussion head injury"' and in other models of animal head injury.'" The levels of hypertension and hypotension are both highly related to the severity of the injury. However, at high levels (> 3.0 atm) the structural insult is more complex and the correlation with blood pressure alterations is less clear."'' Moreover, the bimodal response at this level of injury indicates that a 3.0-atm impact appears to be a critical threshold for the cat fluid-percussion model and the level at which significant brain-stem injury is produced. The observation that both mild and severe injury was observed within a narrow percussion severity range indicates that structural failure of the brain stem occurs rapidly and in a nongraded fashion. It is interesting that the animals with mild brain-stem injury recovered from hypotension within 2 hours, while hypotension in animals with severe brain-stem injury was prolonged and did not resolve. One might infer that the degree and rate of recovery is related to the magnitude of autoregulatory dysfunction of the sympathetic cardiovascular center located in the brain stem.' This appears to be a graded response. With less structural damage (Group 1), 2 hours were required far complete recovery from hypotension and return to baseline. Alterations in ICP and PH Narayan, et al., 28 reported that the ICP in combination with multimodality evoked potentials may be the two strongest predictors of the outcome following severe head injury. Our studies show that the ICP course immediately following brain-stem injury may be deceptive. Despite the marked differences in tissue damage between groups, the ICP rise measured at 8 hours postinjury in the more-severely injured animals averaged 19 mm Hg compared to 14 mm Hg in Group 1 cats. This difference in ICP levels measured at 8 hours did not reach statistical significance. However, the PVI's of severely injured animals were clearly lower than those in Group 1 (p < 0.05), indicating that brain swelling was increased in proportion to severity of injury. These results are similar to clinical observations' and provide further support for the notion that the .1. Neurosurg. / Volume 74 / February, 1991
PVI value is a more sensitive prognostic indicator than ICP during the acute stage of severe head injury. Effect ref Cerebral Blood Plow Changes in local CBF after head injury in humans have been well studied: 4 ' 17 ' 7. " and reductions of CBF below ischemic thresholds have been shown to be related to outcome after head injury. However, in experimental studies, reported changes in local CBF reveal wide variations."'" Although we measured the CBF only in the cortex, the cortical CBF values paralleled those of the physiological parameters as noted above. The previous reports describing CBF response following fluid-percussion trauma focused on the CBF changes immediately following injury."' DeWitt, et al.,' used the microsphere technique and reported that CBF evaluated 1 minute after both mild and high (average 2.68 atm) levels of central injury returned to preinjury level by 60 minutes. Hyperemia is commonly observed in the first few moments after fluid percussion at lower injury levels followed by a gradual tendency toward the baseline value.' Our data indicate that, at higher levels of percussion sufficient to produce brain-stem injury, the CBF is significantly decreased in each hemisphere of both groups at 2 hours posttrauma. Although the CBF in Group 1 animals returned to 89% of the baseline value, the CBF in Group 2 cats remained depressed and did not recover in either hemisphere. These observations agree with the studies of Heiss and Dellinger," who reported that the degree of brain-stem injury correlated highly with the rate of CBF reduction in patients comatose after head trauma. The close association between changes in cortical CBF and the degree of brain-stem injury following severe head trauma suggests that the brain stem plays a part in CBF control. Hawkins, et al.,' observed that bilateral lesions in the mesencephalic reticular formation reduced CBF markedly in all gray and white matter lesions at 6 hours. In addition, the posttraumatic hypotension developed in this model may also reflect damage to brain-stern centers augmenting cardiovascular control. Adams, et al.,' found that primary or pure brainstem injury was rare in severe head injury. In our study, six of the seven animals with severe brain-stem injury also had multiple hemorrhages in the white matter of both hemispheres similar to the diffuse axonal injury demonstrated by Adams, et al. Despite the moderate supratentorial tissue damage, this experimental model at fatal levels of percussion causes a variety of irreversible neurophysiological alterations in the brain stem. Our temporal observations of CBF and brain-stem evoked potentials suggest that in this model, a period 2 hours postinjury represents a critical turning point. Overgaard and Tweed" found that 50% of the patients with impaired autoregulation within 24 hours after head injury showed a good recovery, and the fatal value of CBF, judged by the xenon intra-arterial injection method, was less than 20 m1/100 gm/min. As shown in 275
K. Shima and A. Marmarou Fig. 4, the borderline between each group in this study was also about 20 m1/100 gm/min. Conclusions The fluid-percussion model at high injury levels is a model of brain-stem injury and the outcome is related to the degree of brain-stern damage. Brain-stem potentials, CBF, and PVI showed progressive deterioration in proportion to structural damage and were correlated with outcome. In the animals that recovered, the gradual return of brain-stem potentials and CBF was initiated 2 hours after insult. This may be the critical turning point in this model at high levels of percussion. Experimental studies of head injury must take into consideration the predominant brain-stem focus of injury at these levels of fluid percussion. It is fair to state that a mechanical model of severe head injury which mimics the supratentorial damage observed in the clinical setting is currently not available. Acknowledgments
We acknowledge Dr. Yamamoto for technical assistance with the neurophysiological measurements and Jana Dunbar for laboratory expertise. We also thank Paula Harlfinger and Martha van den Brink for their assistance in the preparation of this manuscript. References
1. Achor U, Starr A: Auditory brain stem response in the cat. 1. Intracranial and extracranial recordings. Electroencephalogr Clin Neurophysiol 48:154-173,1980
2. Achor LJ, Starr A: Auditory brain stem response in the cat. II. Effects of lesions. Electroencephalogr Clin Neurophysiol 48:174-190, 1980
3. Adams JH, Graham DI, Murray LS, et al: Diffuse axonal injury due to non-missile head injury in humans: an analysis of 45 cases. Ann Neural 12:557-563, 1982 4. Alexander RS: Tonic and reflex functions of medullary sympathetic cardiovascular centers. J Neurophysiol 9: 205-217,1946 5. Anderson DC, Bundlie 5, Rochswold GL: Multimodality evoked potentials in closed head trauma. Arch Neural 41: 369-374,1984 6. Buchwald JS, Huang CM: Far-field acoustic response: origins in the cat. Science 189:382-384, 1975 7. Crockard HA, Brown FD, Johns LM, et al: An experimental cerebral missile injury model in primates. J Neurosurg 46:776-783, 1977 8. DeWitt DS, Jenkins LW, Wei EP, et al: Effects of fluidpercussion brain injury on regional cerebral blood flow and pial arteriolar diameter. J Neurosurg 64:787-794, 1986 9. Faught E, Oh SJ: Brainstem auditory evoked responses in brainstem infarction. Stroke 16:701-705, 1985 10. Fieschi C, Battistini N, Beduschi A, et al: Regional cerebral blood flow and intraventricular pressure in acute head injuries. J Neural Neurosurg Psychiatry 37: 1378-1388,1974 11, Gennarelli TA, Segawa H, Wald U, et al: Physiological response to angular acceleration of the head, in Grossman RG, Gildenberg PL (eds): Head Injury: Basic and Clinical Aspects. New York: Raven Press, 1982, pp 129-140
276
12 Greenberg RP, Becker DP, Miller JD, et al: Evaluation of brain function in severe human head trauma with multimodality evoked potentials. Part 2: Localization of brain dysfunction and correlation with posttraumatic neurological conditions. J Neurosurg 47:163-177, 1977 13. Greenberg RP, Mayer DJ, Becker DP, et al: Evaluation of brain function in severe human head trauma with multimodality evoked potentials. Part 1: Evoked braininjury potentials, methods, and analysis. J Neurosurg 47: 150-162,1977 14. Greenberg RP, Miller JD, Becker DP: Clinical findings associated with brainstem dysfunction; an electrophysiological study in severe human head trauma, in Popp AJ, Bourke RS, Nelson LR, et al (eds): Neural Trauma. New York: Raven Press, 1979, pp 229-236 15. Greenberg RP, Newlon PG, Hyatt MS, et al: Prognostic implications of early multimodality evoked potentials in severe head-injury patients. A prospective study. J Neurosurg 55:227-236, 1981 16. Hawkins RA, Hass WK, Ransohoff J: Cerebral blood flow, glucose utilization, oxidative metabolism, and plasticity after mesencephalic reticular formation lesions, in Popp AJ, Bourke RS, Nelson LR, et al (eds): Neural Trauma. New York: Raven Press, 1979, pp 9-18 17. Heiss WD, Jellinger K: Cerebral blood flow and brainstem lesions. J Neural 203:197-209, 1972 18. Lewelt W, Jenkins LW, Miller JD: Autoregulation of cerebral blood flow after experimental fluid-percussion injury of the brain. J Neurosurg 53:500-511,1980 19. Lewelt W, Jenkins LW, Miller JD: Effects of experimental fluid-percussion injury of the brain on cerebrovascular reactivity to hypoxia and to hypercapnia. J Neurosurg 56:332-338, 1982 20. Lindsay KW, Carlin J, Kennedy I, et al: Evoked potentials in severe head injury — analysis and relation to outcome. J Neural Neurosurg Psychiatry 44:796-802, 1981 21. Marmarou A, Nakamura T, Sakamoto H, et al: Development of brain edema following fluid percussion injury, in Inaba Y, Klatzo I, Spatz M (eds): Brain Edema. New York: Springer-Verlag, 1984, pp 88-91 22. Marmarou A, Shima Comparative studies of edema produced by fluid percussion injury with lateral and central modes of injury in cats, in Long DM (ed): Advances in Neurology, Vol 52. Brain Edema. Pathogenesis, Imaging, and Therapy. New York: Raven Press, 1990, pp
233-236 23. Marmarou A, Shulman K, LaMorgese J: Compartmental analysis of compliance and outflow resistance of the cerebrospinal fluid system, J Neurosurg 43:523-5 34, 1975 24. Marmarou A, Takagi H, Shulman K: Biomechanics of brain edema and effects of local cerebral blood flow, in Cervels-Navarro J, Ferszt R (eds): Brain Edema. New York: Raven Press, 1980, pp 345-358 25. Maset AL, Marmarou A, Ward JD, et al: Pressure-volume index in head injury. J Neurosurg 67:832-840, 1987 26. Meyer JS, Rondo A, Nomura F, et al: Cerebral hemodynamics and metabolism following experimental head injury. J Neurosurg 32:304-319, 1970 27. Muizelaar JP, Lutz HA III, Becker DP: Effect of mannitol on ICP and CBF and correlation with pressure autoregulation in severely head-injured patients. J Neurosurg 61: 700-706,1984 28. Narayan RK, Greenberg RP, Miller JD, et al: Improved confidence of outcome prediction in severe head injury. A comparative analysis of the clinical examination, multimodality evoked potentials, CT scanning, and intracranial pressure. J Neurosurg 54:751-762, 1981 J. Neurosurg. / Volume 74 / February, 1991
Brain-stem dysfunction after fluid-percussion head injury 29. Obrist WD, Langfitt TW, Jaggi JL, et al: Cerebral blood flow and metabolism in comatose patients with acute head injury. Relationship to intracranial hypertension. J Neurosurg 61:241-253,1984 30. Overgaard J, Tweed WA: Cerebral circa elation after head injury. Part 1: Cerebral blood flow and its regulation after closed head injury with emphasis on clinical correlations. J Neurosurg 41:531-541, 1974 31. Rosenberg C, Wogensen K, Starr A: Auditory brain-stem and middle and long-latency evoked potentials in coma. Arch Neurol 41:835-838, 1980 32. Rosner MJ, Bennett MD, Becker DP: The clinical relevance of laboratory head injury models: prerequisites for therapeutic testing, in Grossman RG, Gildenberg PL (eds): Head Injury: Basic and Clinical Aspects. New York: Raven Press, 1982, pp 103-116 33. Rosner MJ, Newsome HH, Becker DP: Mechanical brain injury: the sympathoadrenal response. J Neurosurg 61: 76-86,1984 34. Sullivan HG, Martinez J, Becker DP, et al: Fluid-percussion model of mechanical brain injury in the cat. J Neurosurg 45:520-534, 1976
J. Neurosurg. / Volume 74 / February, 1991
35. Tsubokawa T. Nishimoto H, Yamamoto T. et al: Assessment of brainstem damage by the auditory brainstem response in acute severe head injury. J Neurol Neurosurg Psychiatry 43:1005-101 1, 1980 36. Unterberg AW, Andersen BJ, Clarke GD, et al: Cerebral energy metabolism following fluid-percussion brain injury in cats. J Neurosurg 68:594-600,1988 37. Uziel A, Benezech J: Auditory brain-stem responses in comatose patients: relationship with brain stem reflexes and levels of coma. Electroencephalogr Clin Neurophysiol 45:515-524.1978
Manuscript received February 2,1990. Accepted in final form June 18,1990. This work was supported in part by Grants NS 12587 and NS 19235 from the National Institutes of Health. Additional facilities and support were provided by the Richard Roland Reynolds Neurosurgical Research Laboratories. Address reprint requests to. Anthony Marmarou, Ph.D., Division of Neurosurgery, Medical College of Virginia, MCV Station Box 508, Richmond, Virginia 23298. -
277
Evaluation of brain-stem dysfunction following severe fluid-percussion head injury to the cat KATSUJI SHIN4A, M.D., D.M.Sc., AND ANTHONY MARMAROU, PH.D. Division of Neurosurgery, Medical College of Virginia, Richmond, Virginia v The degree of brain-stem dysfunction associated with high-level fluid-percussion injury (3.0 to 3.8 atm) was investigated in anesthetized cats. Measurements were made of the animals' intracranial pressure (ICP), pressure-volume index (FYI), far-field brain-stem auditory evoked responses (BAER's), and cerebral blood flow (CBF). The animals were classified into two groups based on the severity of neuropathological damage to the brain stem after trauma: Group 1 had mild intraparenchymal and subarachnoid hemorrhages and Group 2 had severe intraparenchymal and subarachnoid hemorrhages. The ICP values in Group 1 were insignificantly lower than those in Group 2, while the PVI values in Group 2 were clearly lower (p < 0.05). Immediately after the injury, peaks H, III, and IV of the BAER's demonstrated a transitory and marked suppression. One Group 1 and two Group 2 animals showed the disappearance of peak V. In Group 1, the latencies of peak II, III and IV gradually increased until 60 to 150 minutes postinjury, then returned to 95% of baseline value at 8 hours; however, the animals in Group 2 showed poor recovery of latencies. Two hours after brain injury, the CBF decreased to 40% of the preinjury measurement in both groups (p < 0.001). In contrast to Group 2, the CBF in Group I returned to 86.8% of the preinjury measurement by 8 hours following the injury. Changes in PVI, BAER, and CBF correlated well with the degree of brain-stem injury following severe head injury. These data indicate that high-level fluid-percussion injury (> 3.0 atm) is predominantly a model of brain-stem injury. ,
KEY WORDS • head injury • pressure-volume index • fluid-percussion injury • brain-stem auditory evoked response • cerebral blood flow • cat
T
model has been used widely in the production of experimental head injury. Investigators using this model have made substantial contributions to the body of knowledge elucidating mechanisms of brain trauma. In recent studies, conducted at high injury levels, we observed that the edema resulting from severe experimental brain injury in animals that died shortly after injury was maximal in brain-stem areas.' We hypothesized that animals subjected to high levels of fluid percussion, which involves significant fluid volume injection into the closed calvaria, may be predisposed to severe brain-stem damage when tissue is displaced and the structural threshold is exceeded. This might explain the bimodel distribution in mortality seen frequently in animals subjected to high levels of fluid-percussion injury. The evidence from clinical" .12 . 35 and experimental' 26.34 neuropathology indicates that the degree of brainstem damage following severe head trauma is one of the most important factors influencing outcome. Adams, et al., 3 reported that the rostral aspect of the brain 270
HE fluid-percussion
stem is a common site for the most severe degree of diffuse axonal injury and of multiple petechial hemorrhaging, which is found frequently in patients who die very shortly after head trauma. Heretofore, the characteristics of the fluid-percussion model at high levels of injury have not been addressed. Knowledge of both the mechanical behavior and pathology produced at high injury levels is essential to the study of lethal percussion injury and the similarity of experimental and clinical observations. Thus, the first objective of the present study was to detect and quantitatively describe the severity of brain-stem damage following high-level fluid-percussion injury utilizing both neurophysiological analysis with the brain-stem auditory evoked response (BAER) and brain compliance with the pressure-volume index (PVI). These measurements have been shown to be sensitive indicators of brain-stern dysfunction and brain swelling. 5. 35 9,12-15 - 2 °. ." With the pathophysiological response to severe fluid percussion defined, the second objective was to study the structural response. This was accomplished J. Neurosurg_ / Volume 74 /February, 1991
Brain-stem dysfunction after fluid-percussion head i njury in collaboration with the bioengineering laboratories of the University of Pennsylvania and will be described in a separate report.
Materials and Methods Adult cats, weighing between 2.5 and 4.0 kg, were anesthetized using intravenous rnethohexital sodium titrated to the absence of blink reflex (2 to 5 ml of a 1% solution). Following initial induction of anesthesia, the animals were ventilated with a gas mixture containing 70% NO and 30% 0 2 coupled with a supplemental barbiturate (35 mg/kg) sufficient to maintain surgical depth of anesthesia for cannulation of the femoral artery and vein and cranial preparation for the injury coupling device. During these procedures, gallamine triethiodide (4 mg/kg) was used to maintain muscle relaxation for mechanical ventilation and a reliable endtidal CO 2 concentration of 4% to 4.5% (arterial pCO 2 between 26 and 32 mm Hg). During the entire experimental period, adequate depth of anesthesia was assessed by observing the blood pressure and heart rate response to paw pinch. A heating pad was used to maintain body temperature at 37° to 39°C. Following these procedures, the animals were placed on a stereotactic frame and maintained in a sphinx position with hollow ear bars, avoiding damage to the auditory canal that might interfere with accurate BAER measurement. After the scalp was reflected, a craniotomy 11 mm in diameter was made over the left temporal lobe just rostra] to the external auditory meatus, and a hollow metal screw was tightly inserted in the cranium over the intact dura. Bilateral small craniotomies, 3 mm in diameter, were made in the frontal bone prior to brain injury and in the parietal bone immediately following the injury for insertion of Teflon-coated platinum electrodes for the measurement of cerebral blood flow (CBF). A stainless-steel screw was secured at the vertex for measuring far-field BAER's, and a second reference electrode was inserted into the inner pinna of the right ear. Bifrontal screws for electroencephalography (EEG) were also affixed to the bone. An additional two electrodes at the brow and the occipital protuberance served as grounds for BAER measurement and EEG, respectively.
Measurement of BAER The BAER's were recorded by means of an averager.* Stimulation was by 100 Asec square-wave clicks delivered at 10.17sec and 75 dB sound pressure level intensity through the polyethylene (PE) tubing (12 cm) connecting the earphone to the external canal. The positive input was obtained from the vertex. Recording electrodes were connected to a preamplifier (150 to 3000 Hz bandpass, 104 gain). The amplified responses were * Averager, Model CA 1000, manufactured by Nicolet Instrument Corp., Madison, Wisconsin.
Neurosurg. / Volume 74 / February, 1991
averaged 500 times at the analysis time of 10 cosec, and charted on an x-y recorder.
Measurement of Local CBF Local CBF was measured with the hydrogen-clearance technique.' Teflon-coated platinum needles, 254 /./ in diameter, were implanted under the operating microscope into the gray matter of the right and left suprasylvian gyrus according to stereotactic coordinates (preinjury: +16 AP, 8.5 Lat; postinjury: —4 AP, 7.5 Lat). The reference was placed in the nuchal muscle. Hydrogen gas was administered into the inspired gas mixture and adjusted to the constant flow rate of 2%. The electrodes were connected to electrode current amplifiers and the output was then routed to a computer for on-line computation of CBF. The first measurement of CBF was not made until 1 hour after initial electrode placement in order to permit full polarization and stabilization of the electrodes.
Measurement of ICP and PVI Following the injury, the cisterna magna was cannulated with a No. 21 needle with an interposed fourway stopcock array which allowed continuous monitoring of intracranial pressure (ICP) and bolus injection for PVI. The needle was connected to a Statham pressure transducer (P23ID) through PE-50 tubing. Mock cerebrospinal fluid (0.1 ml) was injected into the cisterna magna for the determination of PVI to assess the changes in tissue compliance.'
Experimental Protocol After surgical preparation and before brain injury, the cats were removed from the stereotactic frames and baseline local CBF and BAER were measured. Brain injury was produced by a fluid-percussion device' attached securely to a right-angled hollow screw positioned in the temporal region to produce a lateral injury. High percussion levels between 3.0 and 3.8 atm were selected to produce severe brain injury, and the level of injury for each animal was measured from a pressure pulse recorded on an oscilloscope. Immediately following the injury, the injury tube was removed, and the craniotomy defect was first tilled with Gelfoam then sealed with dental cement. The BAER's were recorded at approximately 5-minute intervals until 30 minutes had passed, and then at 30-minute intervals until the end of each experiment. Local CBF was measured every 1 or 2 hours, while FYI was measured at 30-minute intervals beginning 2 hours after the brain injury. Thirty minutes before sacrifice by intravenous injection of KC1, each cat was injected with a 2% filtrated solution of Evans blue dye. At 15 minutes before sacrifice, the animals were injected with sodium pentobarbital (100 mg/kg) to anesthetize the animals deeply prior to ICI injection. Following sacrifice, macroscopical pathological analysis was carried out in all animals to assess the location and degree of brain injury and Evans blue staining. 271
K. Shima and A. Marmarou
FIG. l . The time course of mean arterial blood pressure (MABP, upper), pressure-volume index (PVI, center), and intracranial pressure (ICP, lower) in five cats with mild brainstem injury (Group 1, left) and in seven cats with severe brainstem injury (Group 2, right) following severe brain injury. Values are mean ± standard deviation. Type = animal group.
FIG. 2. Percent changes in amplitude of each peak of the brain-stem auditory evoked responses (BAER's) in Group 1 (lei) and Group 2 (right) animals following severe head injury. Roman numerals denote peaks in the brain-stem auditory evoked response.
Results
averaged 17.1 ± 3.64 mm Hg and increased mildly to an average of 19.8 ± 6.11 mm Hg at 8 hours. The ICP profiles in these two groups were not significantly different (Fig. 1). Although the ICP profiles among Group 1 and 2 animals were similar, the degree of brain swelling as indexed by PVI was significantly different. The PVI's in Group 1 averaged 0.978 ± 0.132 ml at 2 hours postinjury and 0.756 ± 0.033 ml at 8 hours (Fig. 1). However, in Group 2 animals, the initial PVI's measured at 2 hours were lower, averaging 0.615 ± 0.075 ml, then decreased to 0.498 ± 0.065 ml at 8 hours postinj ury.
The animals were classified into two groups based on the neuropathological changes in the brain stem. Group 1 (five cats) exhibited spot-like or small (< 3 aim in diameter) isolated intraparenchymal hemorrhages and subarachnoid hemorrhages, and Group 2 (seven cats) had diffuse or large (> 3 mm in diameter) multiple intraparenchymal hemorrhages and subarachnoid hemorrhages. The mean percussion injury levels (± standard deviation) were 3.2 ± 0.19 atm for Group 1 and 3.3 ± 0.34 atm for Group 2.
Physiological Changes In the 1st minute after injury, a loss of the corneal and pupillary reflex and a rapid increase in systemic mean arterial blood pressure (MABP) was observed. At 1 minute following the injury, the MABP was significantly increased by 97.0% in Group 1 and 81.6% in Group 2 over each baseline value (p < 0.001); however, there was no significant difference in peak MABP between Groups 1 and 2 (Fig. 1). After the initial rise, the MABP in Group I fell below baseline, then rapidly increased, reaching preinjury levels by 2 hours postimpact. Following these transient changes, the MABP remained at baseline throughout the 8 hours of observation. The initial transient changes of MABP in Group 2 animals were similar to those in Group 1 with the exception that, after crossing baseline, MABP continued to decrease gradually to baseline without recovery (Fig. 1 right). One animal in Group 2 was sacrificed 3.4 hours after injury because of a rapid decrease in MABP.
ICP and PT/1 Responses to Injury Cannulation of the cisterna magna was possible at 2 hours postinjury for measurement of ICP and PVI. In Group 1 animals, the initial ICP averaged 14.25 ± 2.86 mm Hg and remained at that level for the duration of the experiment. In Group 2 animals, the initial ICP 272
Brain-Stern Auditory Evoked Response The first five positive peaks of the BAER's were labeled sequentially from peak I to peak V. Immediately following brain injury, peaks II, III, and IV demonstrated a transient and marked suppression. The amplitude of peak I remained suppressed by 50% in each group except for one Group 2 cat at 2 hours after trauma (Fig. 2). In contrast, one Group 1 and all Group 2 cats showed the disappearance of peak V. In one Group 2 cat peaks III and IV also disappeared during the 8 hours after the injury; however, the amplitude of peak V in Group I animals recovered to 58.4% of the baseline value by 8 hours after trauma, which was close to the value of peak I at 8 hours. Peaks II, III, and IV in Group 1 cats remained at a milder reduction of amplitude than those in Group 2 throughout the experiment. All peaks in Group 1 animals showed a tendency to recover from the lowest values at 4 or 5 hours after the injury. In Group 2, however, each peak was markedly suppressed at a constant level below 40%. Individual examples of the effects of trauma upon the BAER's shown in Fig. 3 are characteristic of each group. The mean baseline latency (± standard deviation) of each peak and range was as follows: peak I, 1.86 ± 0.24 msec; peak II, 2.78 ± 0.17 msec; peak HI, 3.51 ± 0.17 J. Neurosurg. / Volume 74 / February, 1991
Brain-stem dysfunction after fluid-percussion head injury
FIG. 3. Examples of brain-stem auditory evoked responses (BAER's). Roman numerals denote BAER peaks. Left: Two examples of BAER's in Group 1 animals which recovered by 8 hours postinjury. The amplitude and latencies of all waves are not seriously affected. Right: Two examples of BAER's in Group 2 animals showing disappearance of early waves and specifically wave V.
msec; peak IV, 4.43 ± 0.18 msec; and peak V, 5.70 ± 0.39 msec (Fig. 4). In Group 1, the latency of peaks II, III, and IV was prolonged gradually until 60 to 150 minutes postinjury, then returned to approximately 95 % of the baseline value by 8 hours. In contrast, the animals in Group 2 did not show recovery of the latencies of all peaks.
Local Cerebral Blood Flow The preinjury local CBF values in the frontal cortex were 43.3 ± 3.3 m1/100 gm/min (4.2 ± 0.66 m1/100 gm/min in Group 1 cats and 42.7 ± 3.1 ml/gm/min in Group 2 cats). Two hours after the injury, the CBF significantly decreased to approximately 40% from the preinjury levels in each hemisphere in both groups (p < 0.001). The CBF in Group 1 animals gradually returned to 86.8% of the preinjury level by 8 hours
FIG. 4. Sequential changes in the latencies of the brainstem auditory evoked responses (BAER's) in five Group 1 (left) and seven Group 2 (right) animals following severe head injury. Roman numerals denote BAER peaks. Values are mean ± standard deviation. Wave V was absent following injury. Type = animal group.
J. Neurosurg. / Volume 74 / February, 1991
following trauma (Fig. 5). However, the low CBF in Group 2 cats persisted in each hemisphere until the end of the experiment.
Neuropathological Findings The only extravasation of Evans blue dye in the brains of these animals was a heterogeneous leakage in the cerebral cortex under the fluid-percussion tube. All of the cats showed moderate to severe subarachnoid hemorrhage. By classification, the five Group I cats had spot-like or small (< 3 mm in diameter) multiple intraparenchymal hemorrhages in the midbrain to the medulla oblongata. Six of the seven Group 2 animals had multiple petechial hemorrhages in the midbrain and pons, and even in the white matter of both cerebral
FIG. 5. Cerebral blood flow (CBF) in the left (injured) cerebral cortex and the right cerebral cortex following severe head injury. Circles indicate the mean CBF in four Group 1 animals; squares indicate the mean CBF in four Group 2 animals. Values are given ± standard deviation. * = p < 0.05; ** = p < 0.01; *** = p < 0.001; + = p < 0.05 vs. Group 1.
273
K. Shima and A. Marmarou
FIG. 7. Mean sequential changes in interpeak latencies (IPL) of brain-stem auditory evoked response peaks I to III (squares), III to V (circles), and I to V (triangles) in five Group 1 animals. Values are given ± standard deviation.
FIG. 6. Spatial distribution of brain-stem hemorrhages (> 1 mm) in five Group I (upper) and seven Group 2 (lower) animals following severe head injury.
hemispheres (Fig. 6). During the experimental period
there were no significant changes in p0 2 , pCO 2 , or pH. Discussion The results of our studies indicate that high-level fluid percussion is predominantly a brain-stem injury. The neurophysiological changes, specifically the absence of BAER wave V, points to structural damage of the midbrain and upper pons, which is in agreement with the morphological findings. Permanent damage, as assessed by failure of wave V to return, was correlated with sustained hypotension, low brain compliance and reduced CBF. With less tissue damage, both latency and amplitude of the BAER's recovered by 8 hours postinjury. Neurophysiological Correlates: Brain-Stem Injury
Many investigators have attempted to establish a relationship between the changes in the BAER's and brain-stem lesions confirmed by neurological and radiological findings. ' 2 . 14 . 35 ' 37 However, few attempts have been made to evaluate the sequential changes in brainstem dysfunction immediately following severe brain injury. In the current studies, although all BAER peaks demonstrated prolonged latencies and suppressed amplitudes following trauma, these alterations varied for each peak. Greenberg, et a1., 12 established a grading system of abnormal BAER's in patients with severe head trauma. Based on the degree of BAER abnormality, they noted delays of interpeak latency between peaks I and V in 17 of 21 patients. In their Grade III BAER, considered severely abnormal, only peak I has a normal latency and amplitude. Tsubokawa, et al.,' analyzed the BAER's in 64 274
patients within 3 days following severe head injury and divided the cases into two groups. One group had a prolonged latency or the disappearance of peak V, only six of the 13 patients in this group had a good outcome. In the present experimental study, the disappearance of peak V was observed in 10 of 12 cats within 3 minutes posttrauma. Wave V disappeared and did not recover in any cat with greater tissue damage as assessed by an increased number of petechial hemorrhages (Group 2). The absence of peak V suggests damage to the inferior colliculus which is thought to be the generating locus of peak By contrast, peak V in animals with less tissue damage (Group 1) appeared by approximately 20 minutes after injury and interpeak latencies of peaks I to V and peaks III to V, which are an index of central conduction time, returned to the baseline values by 8 hours (Fig. 7), It is of interest that the I to III interpeak latency did not show any change throughout the experiment. Anderson, et al.,' who viewed the I to V interpeak latency as a central auditory conduction time, reported that of 23 surviving patients with severe head trauma nine had the prolongation of I to V interpeak latencies. In our animals, the I to III interpeak latencies did not change during the prolongation of the I to V and II to V interpeak latencies. By contrast, in the patients with brain-stem stroke, the prolongation of the to III interpeak latencies was seen with normal III to V interpeak latencies in the patients with pontomedullary junction infarctions.' At high levels of injury, interpeak latency measures are of limited use since the BAER is obliterated. Thus, the use of latencies between peaks I and III and between peaks III and V is limited to evaluating brain-stem dysfunction and localizing lesions in the acute stage."' Survival With Experimental Brain Stem Injury
Following severe brain injury, the brain stem can be affected directly by a variety of hemorrhages, and later secondarily by uncal herniation. We have recently observed that cats, which were monitored for a 24-hour period following temporal and central fluid-percussion J. Neurosurg. / Volume 74 / February, 1991
Brain-stem dysfunction after fluid-percussion head injury trauma at the level of 2.8 atm, were classified into two groups based upon length of survival. Mean survival durations of the short-term survivors after central (impact on the vertex) and temporal (impact site over the posterior ectosylvian gyrus) impacts were 10.8 and 12.5 hours, respectively: rapid irreversible hypotension began after about 7 hours postinjury in both temporal and central injury.' Moreover, survival time was diminished in those animals with greater reduction in PVI. We suspect that brain-stem dysfunction played an important role in the outcome of these experiments, as the posttraumatic edema was maximal in brain-stern regions. Blood Pressure Alterations Transient hypertension followed by hypotension immediately after brain injury has been well documented in fluid-percussion head injury"' and in other models of animal head injury.'" The levels of hypertension and hypotension are both highly related to the severity of the injury. However, at high levels (> 3.0 atm) the structural insult is more complex and the correlation with blood pressure alterations is less clear."'' Moreover, the bimodal response at this level of injury indicates that a 3.0-atm impact appears to be a critical threshold for the cat fluid-percussion model and the level at which significant brain-stem injury is produced. The observation that both mild and severe injury was observed within a narrow percussion severity range indicates that structural failure of the brain stem occurs rapidly and in a nongraded fashion. It is interesting that the animals with mild brain-stem injury recovered from hypotension within 2 hours, while hypotension in animals with severe brain-stem injury was prolonged and did not resolve. One might infer that the degree and rate of recovery is related to the magnitude of autoregulatory dysfunction of the sympathetic cardiovascular center located in the brain stem.' This appears to be a graded response. With less structural damage (Group 1), 2 hours were required far complete recovery from hypotension and return to baseline. Alterations in ICP and PH Narayan, et al., 28 reported that the ICP in combination with multimodality evoked potentials may be the two strongest predictors of the outcome following severe head injury. Our studies show that the ICP course immediately following brain-stem injury may be deceptive. Despite the marked differences in tissue damage between groups, the ICP rise measured at 8 hours postinjury in the more-severely injured animals averaged 19 mm Hg compared to 14 mm Hg in Group 1 cats. This difference in ICP levels measured at 8 hours did not reach statistical significance. However, the PVI's of severely injured animals were clearly lower than those in Group 1 (p < 0.05), indicating that brain swelling was increased in proportion to severity of injury. These results are similar to clinical observations' and provide further support for the notion that the .1. Neurosurg. / Volume 74 / February, 1991
PVI value is a more sensitive prognostic indicator than ICP during the acute stage of severe head injury. Effect ref Cerebral Blood Plow Changes in local CBF after head injury in humans have been well studied: 4 ' 17 ' 7. " and reductions of CBF below ischemic thresholds have been shown to be related to outcome after head injury. However, in experimental studies, reported changes in local CBF reveal wide variations."'" Although we measured the CBF only in the cortex, the cortical CBF values paralleled those of the physiological parameters as noted above. The previous reports describing CBF response following fluid-percussion trauma focused on the CBF changes immediately following injury."' DeWitt, et al.,' used the microsphere technique and reported that CBF evaluated 1 minute after both mild and high (average 2.68 atm) levels of central injury returned to preinjury level by 60 minutes. Hyperemia is commonly observed in the first few moments after fluid percussion at lower injury levels followed by a gradual tendency toward the baseline value.' Our data indicate that, at higher levels of percussion sufficient to produce brain-stem injury, the CBF is significantly decreased in each hemisphere of both groups at 2 hours posttrauma. Although the CBF in Group 1 animals returned to 89% of the baseline value, the CBF in Group 2 cats remained depressed and did not recover in either hemisphere. These observations agree with the studies of Heiss and Dellinger," who reported that the degree of brain-stem injury correlated highly with the rate of CBF reduction in patients comatose after head trauma. The close association between changes in cortical CBF and the degree of brain-stem injury following severe head trauma suggests that the brain stem plays a part in CBF control. Hawkins, et al.,' observed that bilateral lesions in the mesencephalic reticular formation reduced CBF markedly in all gray and white matter lesions at 6 hours. In addition, the posttraumatic hypotension developed in this model may also reflect damage to brain-stern centers augmenting cardiovascular control. Adams, et al.,' found that primary or pure brainstem injury was rare in severe head injury. In our study, six of the seven animals with severe brain-stem injury also had multiple hemorrhages in the white matter of both hemispheres similar to the diffuse axonal injury demonstrated by Adams, et al. Despite the moderate supratentorial tissue damage, this experimental model at fatal levels of percussion causes a variety of irreversible neurophysiological alterations in the brain stem. Our temporal observations of CBF and brain-stem evoked potentials suggest that in this model, a period 2 hours postinjury represents a critical turning point. Overgaard and Tweed" found that 50% of the patients with impaired autoregulation within 24 hours after head injury showed a good recovery, and the fatal value of CBF, judged by the xenon intra-arterial injection method, was less than 20 m1/100 gm/min. As shown in 275
K. Shima and A. Marmarou Fig. 4, the borderline between each group in this study was also about 20 m1/100 gm/min. Conclusions The fluid-percussion model at high injury levels is a model of brain-stem injury and the outcome is related to the degree of brain-stern damage. Brain-stem potentials, CBF, and PVI showed progressive deterioration in proportion to structural damage and were correlated with outcome. In the animals that recovered, the gradual return of brain-stem potentials and CBF was initiated 2 hours after insult. This may be the critical turning point in this model at high levels of percussion. Experimental studies of head injury must take into consideration the predominant brain-stem focus of injury at these levels of fluid percussion. It is fair to state that a mechanical model of severe head injury which mimics the supratentorial damage observed in the clinical setting is currently not available. Acknowledgments
We acknowledge Dr. Yamamoto for technical assistance with the neurophysiological measurements and Jana Dunbar for laboratory expertise. We also thank Paula Harlfinger and Martha van den Brink for their assistance in the preparation of this manuscript. References
1. Achor U, Starr A: Auditory brain stem response in the cat. 1. Intracranial and extracranial recordings. Electroencephalogr Clin Neurophysiol 48:154-173,1980
2. Achor LJ, Starr A: Auditory brain stem response in the cat. II. Effects of lesions. Electroencephalogr Clin Neurophysiol 48:174-190, 1980
3. Adams JH, Graham DI, Murray LS, et al: Diffuse axonal injury due to non-missile head injury in humans: an analysis of 45 cases. Ann Neural 12:557-563, 1982 4. Alexander RS: Tonic and reflex functions of medullary sympathetic cardiovascular centers. J Neurophysiol 9: 205-217,1946 5. Anderson DC, Bundlie 5, Rochswold GL: Multimodality evoked potentials in closed head trauma. Arch Neural 41: 369-374,1984 6. Buchwald JS, Huang CM: Far-field acoustic response: origins in the cat. Science 189:382-384, 1975 7. Crockard HA, Brown FD, Johns LM, et al: An experimental cerebral missile injury model in primates. J Neurosurg 46:776-783, 1977 8. DeWitt DS, Jenkins LW, Wei EP, et al: Effects of fluidpercussion brain injury on regional cerebral blood flow and pial arteriolar diameter. J Neurosurg 64:787-794, 1986 9. Faught E, Oh SJ: Brainstem auditory evoked responses in brainstem infarction. Stroke 16:701-705, 1985 10. Fieschi C, Battistini N, Beduschi A, et al: Regional cerebral blood flow and intraventricular pressure in acute head injuries. J Neural Neurosurg Psychiatry 37: 1378-1388,1974 11, Gennarelli TA, Segawa H, Wald U, et al: Physiological response to angular acceleration of the head, in Grossman RG, Gildenberg PL (eds): Head Injury: Basic and Clinical Aspects. New York: Raven Press, 1982, pp 129-140
276
12 Greenberg RP, Becker DP, Miller JD, et al: Evaluation of brain function in severe human head trauma with multimodality evoked potentials. Part 2: Localization of brain dysfunction and correlation with posttraumatic neurological conditions. J Neurosurg 47:163-177, 1977 13. Greenberg RP, Mayer DJ, Becker DP, et al: Evaluation of brain function in severe human head trauma with multimodality evoked potentials. Part 1: Evoked braininjury potentials, methods, and analysis. J Neurosurg 47: 150-162,1977 14. Greenberg RP, Miller JD, Becker DP: Clinical findings associated with brainstem dysfunction; an electrophysiological study in severe human head trauma, in Popp AJ, Bourke RS, Nelson LR, et al (eds): Neural Trauma. New York: Raven Press, 1979, pp 229-236 15. Greenberg RP, Newlon PG, Hyatt MS, et al: Prognostic implications of early multimodality evoked potentials in severe head-injury patients. A prospective study. J Neurosurg 55:227-236, 1981 16. Hawkins RA, Hass WK, Ransohoff J: Cerebral blood flow, glucose utilization, oxidative metabolism, and plasticity after mesencephalic reticular formation lesions, in Popp AJ, Bourke RS, Nelson LR, et al (eds): Neural Trauma. New York: Raven Press, 1979, pp 9-18 17. Heiss WD, Jellinger K: Cerebral blood flow and brainstem lesions. J Neural 203:197-209, 1972 18. Lewelt W, Jenkins LW, Miller JD: Autoregulation of cerebral blood flow after experimental fluid-percussion injury of the brain. J Neurosurg 53:500-511,1980 19. Lewelt W, Jenkins LW, Miller JD: Effects of experimental fluid-percussion injury of the brain on cerebrovascular reactivity to hypoxia and to hypercapnia. J Neurosurg 56:332-338, 1982 20. Lindsay KW, Carlin J, Kennedy I, et al: Evoked potentials in severe head injury — analysis and relation to outcome. J Neural Neurosurg Psychiatry 44:796-802, 1981 21. Marmarou A, Nakamura T, Sakamoto H, et al: Development of brain edema following fluid percussion injury, in Inaba Y, Klatzo I, Spatz M (eds): Brain Edema. New York: Springer-Verlag, 1984, pp 88-91 22. Marmarou A, Shima Comparative studies of edema produced by fluid percussion injury with lateral and central modes of injury in cats, in Long DM (ed): Advances in Neurology, Vol 52. Brain Edema. Pathogenesis, Imaging, and Therapy. New York: Raven Press, 1990, pp
233-236 23. Marmarou A, Shulman K, LaMorgese J: Compartmental analysis of compliance and outflow resistance of the cerebrospinal fluid system, J Neurosurg 43:523-5 34, 1975 24. Marmarou A, Takagi H, Shulman K: Biomechanics of brain edema and effects of local cerebral blood flow, in Cervels-Navarro J, Ferszt R (eds): Brain Edema. New York: Raven Press, 1980, pp 345-358 25. Maset AL, Marmarou A, Ward JD, et al: Pressure-volume index in head injury. J Neurosurg 67:832-840, 1987 26. Meyer JS, Rondo A, Nomura F, et al: Cerebral hemodynamics and metabolism following experimental head injury. J Neurosurg 32:304-319, 1970 27. Muizelaar JP, Lutz HA III, Becker DP: Effect of mannitol on ICP and CBF and correlation with pressure autoregulation in severely head-injured patients. J Neurosurg 61: 700-706,1984 28. Narayan RK, Greenberg RP, Miller JD, et al: Improved confidence of outcome prediction in severe head injury. A comparative analysis of the clinical examination, multimodality evoked potentials, CT scanning, and intracranial pressure. J Neurosurg 54:751-762, 1981 J. Neurosurg. / Volume 74 / February, 1991
Brain-stem dysfunction after fluid-percussion head injury 29. Obrist WD, Langfitt TW, Jaggi JL, et al: Cerebral blood flow and metabolism in comatose patients with acute head injury. Relationship to intracranial hypertension. J Neurosurg 61:241-253,1984 30. Overgaard J, Tweed WA: Cerebral circa elation after head injury. Part 1: Cerebral blood flow and its regulation after closed head injury with emphasis on clinical correlations. J Neurosurg 41:531-541, 1974 31. Rosenberg C, Wogensen K, Starr A: Auditory brain-stem and middle and long-latency evoked potentials in coma. Arch Neurol 41:835-838, 1980 32. Rosner MJ, Bennett MD, Becker DP: The clinical relevance of laboratory head injury models: prerequisites for therapeutic testing, in Grossman RG, Gildenberg PL (eds): Head Injury: Basic and Clinical Aspects. New York: Raven Press, 1982, pp 103-116 33. Rosner MJ, Newsome HH, Becker DP: Mechanical brain injury: the sympathoadrenal response. J Neurosurg 61: 76-86,1984 34. Sullivan HG, Martinez J, Becker DP, et al: Fluid-percussion model of mechanical brain injury in the cat. J Neurosurg 45:520-534, 1976
J. Neurosurg. / Volume 74 / February, 1991
35. Tsubokawa T. Nishimoto H, Yamamoto T. et al: Assessment of brainstem damage by the auditory brainstem response in acute severe head injury. J Neurol Neurosurg Psychiatry 43:1005-101 1, 1980 36. Unterberg AW, Andersen BJ, Clarke GD, et al: Cerebral energy metabolism following fluid-percussion brain injury in cats. J Neurosurg 68:594-600,1988 37. Uziel A, Benezech J: Auditory brain-stem responses in comatose patients: relationship with brain stem reflexes and levels of coma. Electroencephalogr Clin Neurophysiol 45:515-524.1978
Manuscript received February 2,1990. Accepted in final form June 18,1990. This work was supported in part by Grants NS 12587 and NS 19235 from the National Institutes of Health. Additional facilities and support were provided by the Richard Roland Reynolds Neurosurgical Research Laboratories. Address reprint requests to. Anthony Marmarou, Ph.D., Division of Neurosurgery, Medical College of Virginia, MCV Station Box 508, Richmond, Virginia 23298. -
277
