Dynamic autoregulatory response after severe head injury

J Neurosurg 97:1054 1061, 2002 Dynamic autoregulatory response after severe head injury ROMAN HLATKY, M.D., YU FURUYA, M.D., PH.D., ALEX B. VALADKA, M.D., JORGE GONZALEZ, M.D., ARI CHACKO, M.D., YASU MIZUTANI, M.D., CHARLES F. CONTANT, PH.D., AND CLAUDIA S. ROBERTSON, M.D., F.C.C.M. Department of Neurosurgery, Baylor College of Medicine, Houston, Texas; Department of Neurosurgery, University Hospital, Charles University, Hradec Kralove, Czech Republic; and Department of Neurosurgery, St. Marianna University School of Medicine, Kawasaki, Japan CEREBROVASCULAR Object. The purpose of this study was to evaluate the extent and timing of impairment of cerebral pressure autoregulation after severe head injury. Methods. In a prospective study of 122 patients with severe head trauma (median Glasgow Coma Scale Score 6), dynamic tests of pressure autoregulation were performed every 12 hours during the first 5 days postinjury and daily during the next 5 days. The autoregulatory index ([ARI] normal value 5 1.1) was calculated for each test. The changes in the ARI over time were examined and compared with other physiological variables. The ARI averaged 2.8 1.9 during the first 12 hours postinjury, and continued to decrease to a nadir of 1.7 1.1 at 36 to 48 hours postinjury. At this nadir, in 87% of the patients the value was less than 2.8. This continued deterioration in the ARI during the first 36 to 48 hours postinjury occurred despite an increase in cerebral blood flow ([CBF], p 0.05) and in middle cerebral artery blood flow velocity ([BFV], p 0.001), and could not be explained by changes in cerebral perfusion pressure, end-tidal CO 2, or cerebral metabolic rate of O 2. A marked decrease in cerebrovascular resistance ([CVR], p 0.001) accompanied this deterioration in the ARI. Patients with a relatively higher BFV on Day 1 had a lower CVR (p 0.05) and more impaired pressure autoregulation than those with a lower BFV. Conclusions. The inability of cerebral vessels to regulate CBF normally may play a role in the vulnerability of the injured brain to secondary ischemic insults. These studies indicate that this vulnerability continues and even increases beyond the first 24 hours postinjury. Local factors affecting cerebrovascular tone may be responsible for these findings. KEY WORDS head injury autoregulation cerebral blood flow cerebrovascular resistance transcranial Doppler ultrasonography abnormalities induced by TBI range from ischemia to hyperemia. 3,15 The severity of the reduction in CBF posttrauma, especially in the first few hours after injury, is associated with outcome, 3,4,12,22,23 and these vascular changes contribute to some of the clinical problems encountered during the initial recovery period from trauma, especially the vulnerability to secondary ischemic insults 9 and intracranial hypertension. 15,21 The mechanisms that the brain normally uses to regulate CBF have been examined in an attempt to understand the pathological processes responsible for these trauma-induced cerebrovascular changes. Cerebral pressure autoregulation is the intrinsic ability of the brain to maintain a constant CBF under conditions Abbreviations used in this paper: ANOVA = analysis of variance; ARI = autoregulatory index; AVDO 2 = arteriovenous oxygen difference; BFV = blood flow velocity; BP = blood pressure; CBF = cerebral blood flow; CMRO 2 = cerebral metabolic rate of O 2 ; CPP = cerebral perfusion pressure; CVR = cerebrovascular resistance; ETCO 2 = end-tidal carbon dioxide; GCS = Glasgow Coma Scale; ICP = intracranial pressure; MABP = mean arterial BP; MCA = middle cerebral artery; SD = standard deviation; SjvO 2 = jugular venous O 2 saturation; TBI = traumatic brain injury; Xe-CT = xenonenhanced computerized tomography. of changing CPP. Mechanisms for cerebral pressure autoregulation include myogenic, neurogenic, and/or metabolic processes. These processes continuously adjust CVR for changes in CPP so that the CBF normally remains relatively constant. The upper and lower limits of autoregulation reflect the points at which vasomotor adjustments are exhausted and CVR can neither increase nor decrease to maintain CBF. Impairment of the pressure autoregulatory response has been reported in patients with both minor and severe TBI and may contribute to their vulnerability to secondary brain injury. 6,7,11,17,19,26 Secondary ischemic insults such as hypotension and hypoxia after TBI have been observed to double the mortality rate, presumably because the injured brain is unable to maintain an adequate CBF during these events. 5,9,16 These observations about impaired pressure autoregulation form the basis for much of the current treatment strategy for severe TBI, especially with regard to management of CPP. 24 In most of the previous studies in which impaired cerebral autoregulation was demonstrated in humans after TBI, static methods were used; these compared steady-state CBF at two levels of CPP. Dynamic procedures for assessing cerebral autoregulation, which are more practical in critically ill patients, have been described recently. 1,11,18,20,28 These 1054

Dynamic autoregulatory response after severe head injury methods have in common the use of transcranial Doppler ultrasonography to measure continuously the BFV in the MCA during transient perturbations in the CPP. In one dynamic procedure, rapid deflation of thigh cuffs is used to induce a transient drop in BP. The rate of recovery of the MCA BFV after this rapid decrease in BP gives an index of dynamic autoregulation. Impairments in dynamic cerebral autoregulation found with this method have been closely correlated with impaired autoregulation measured using static BP changes. 25 The purpose of this study was to use this new procedure for measuring dynamic autoregulation to evaluate the extent and timing of disturbances in cerebral pressure autoregulation in severely head injured patients. Clinical Material and Methods Patient Characteristics One hundred twenty-two patients who had a GCS 27 score of 8 or less on admission or whose GCS score deteriorated to 8 or less within 48 hours of admission were studied between March 1998 and October 2000. The research protocol was approved by the Baylor Institutional Review Board for Human Subject Research, and informed consent for participation in the study was obtained from each patient s nearest relative. When relatives were unavailable to give informed consent, the patients were enrolled in the study according to an approved emergency consent procedure. The demographic characteristics of the patients are summarized in Table 1. General Management Protocol for TBI All patients in the study were treated according to a standard protocol that emphasized prompt evacuation of intracranial mass lesions and prevention of secondary insults to the brain. The ICP was monitored using a ventriculostomy catheter, and ICP values greater than 20 mm Hg were treated according to standard protocols. The SjvO 2 was monitored in the dominant jugular vein by using a fiberoptic O 2 saturation catheter. Systemic factors that exacerbate intracranial hypertension, including hypoxia, hypercapnia, fever, and hypotension, were corrected. Adequate volume resuscitation, proper sedation, and selective use of vasopressors were used to maintain BP. An MABP of at least 80 mm Hg and a CPP of at least 60 mm Hg were maintained, unless a low SjvO 2 indicated that a higher CPP was required. The same standard protocol was used throughout the study. Experimental Measurements Dynamic Testing of Cerebral Pressure Autoregulation. The MCA BFV was monitored bilaterally with a transcranial Doppler ultrasonography system equipped with dual 2-MHz transducers (DWL Multi Flow, Sipplingen, Germany). The BP was monitored through a radial artery catheter. To decrease BP transiently, large BP cuffs were wrapped around each upper thigh, then the cuffs were inflated to 20 to 40 mm Hg above systolic BP for 3 minutes. Rapid deflation of the cuffs resulted in an abrupt transient drop in MABP. The dynamic cerebral autoregulatory response was analyzed by observing the relative changes in the MCA TABLE 1 Demographic characteristics of 122 patients with TBI* Characteristic BFV immediately after the BP drop, as shown in Fig. 1. The result was expressed as the ARI, as described by Aaslid, et al., 1 and Tiecks, et al. 28 The ARI is a 10-step scale ranging from 0 (absent autoregulation) to 9 (best autoregulation). The ARI in healthy individuals averages 5 1.1 (mean SD). For the purposes of this study, values of the ARI lower than 2.8 (mean 2 SDs) were considered clearly abnormal. Dynamic tests of autoregulation were performed every 12 hours during the first 5 days and once per day during the next 5 days after TBI. A minimum of three cuff deflation tests was performed; all test results were averaged. An abrupt decrease in MABP of at least 20 mm Hg was required for a test to be included in the data analysis. No adverse events were recorded arising from the dynamic testing of autoregulation. Measurement of Global CBF Eighty-three measurements of CBF were performed using stable Xe-CT scanning (Diversified Diagnostics Products, Houston, TX) in the patients studied. Fifty-four examinations were performed during the first 24 hours and 29 were performed between 24 and 48 hours after TBI. For each examination, the CBFs in all cortical regions of interest were averaged to estimate a global value for CBF. This value was used for calculation of CVR by the equation CVR = CPP/CBF. The normal value of CVR is 1.6 0.4 mm Hg/ml/100 g/min. 13 The CMRO 2 was calculated by the formula CMRO 2 = AVDO 2 CBF. The normal value of CMRO 2 is 3.3 0.4 ml/100 g/min. 13 For territorial measurements of CBF, cortical regions corresponding to the area supplied by the MCA were averaged. Measurement of Cortical CBF Value age (yrs) 35.7 15.2 sex male 104 female 18 initial GCS score 6 range 3 11 injury type evacuated mass 68 unevacuated mass 8 diffuse injury 1 or 2 20 diffuse injury 3 or 4 26 mechanism of injury motor vehicle accident 59 auto pedestrian accident 19 assault 27 gunshot wound 9 other 8 Abbreviated Injury Scale score 26.3 8.6 GOS score at 6 mos good recovery 18 moderate disability 19 severe disability 28 vegetative 6 dead 25 * GOS = Glasgow Outcome Scale. According to Jennett and Bond; 26 patients were lost to follow up. Forty-six of the 122 patients also underwent intraopera- 1055

R. Hlatky, et al. FIG. 1. Graphs plotting examples of the dynamic autoregulation test. Upper: Changes in MABP (ABP) induced by thigh cuff deflation (arrows). Lower: Changes in MCA BFV on the left (MCA L) and right (MCA R) resulting from thigh cuff deflation. In this example, the response of the right MCA is normal, and the tracing for this artery recovers from the drop in MABP, whereas the tracing for the left MCA does not recover and instead follows the MABP. CCP = critical closing pressure; RoR = rate of recovery. tive placement of a thermal-diffusion CBF probe (Flowtronics, Inc., Phoenix, AZ). The probe was placed on the cortex, and care was taken to avoid large cortical vessels. Attention was paid to be sure that the two gold discs were in contact with the cortex. The dura was closed over the CBF sensor, and the bone flap was replaced. The probe was connected to the CBF monitor and calibrated, and the thermal-diffusion CBF values were collected at 30-second intervals along with the other physiological data. Measurement of Brain Tissue PO 2 The brain tissue PO 2 was continuously measured using a miniaturized Clark electrode-licox probe (GMS, Kiel- Mielkendorf, Germany) in 86 of the 122 patients. An area of brain in the frontotemporal region that appeared to be injured but was not clearly necrotic was targeted for the probe placement. Alongside the brain tissue PO 2 probe, a temperature probe was placed in the cortex, and both were connected to the monitor for automatic temperature-corrected brain tissue PO 2 readings. The position of the brain tissue PO 2 probe was examined on follow-up CT scans. The brain tissue PO 2 values were continuously collected and stored at 30-second intervals in a computer. At the end of the monitoring period, the probes were removed, and calibration drift was determined by measuring a stable PO 2 in a zero O 2 solution, in a Level I arterial blood gas control solution (Ciba Corning Blood Gas Control, Medfield, MA), and in room air. Statistical Analysis All summary data are expressed as the mean SD or as the median (interquartile range). For the summary data, the mean of variables from the entire period of monitoring was calculated for each patient. The mean values were compared using the Student t-test. Comparison of the time course between groups was performed using two-way repeated-measures ANOVA, followed by the Tukey test when multiple comparisons were made. FIG. 2. Graphs showing the time course of changes in the ARI and MCA BFV (MCA FV). Both parameters are expressed as the median and interquartile ranges. Statistical analysis was performed using one-way repeated-measures ANOVA. Results Dynamic Testing of Cerebral Pressure Autoregulation During the first 10 days postinjury, 5430 cuff deflation tests were performed in the 122 patients. A total of 4280 of 1056

Dynamic autoregulatory response after severe head injury these tests (35/patient) met the inclusion criterion of inducing an abrupt drop in MABP of at least 20 mm Hg. The mean drop in MABP induced by cuff deflation was 23.4 2.9 mm Hg. The mean duration of testing was 5.8 2.5 days. For each patient, the mean values for the ARI during each time interval were used in analyses of the time course. Time Course of Cerebral Pressure Autoregulation After Severe TBI Compared with the range for the ARI of 2.8 to 7.2 (mean 5) observed in healthy volunteers, values of the ARI were abnormally low in most of the patients with TBI throughout the entire monitoring period. In addition, there was a characteristic evolution of the ARI over time in the individual patients with TBI. This typical pattern consisted of a low ARI on admission that continued to decrease throughout the first 36 hours postinjury. After the nadir at 36 to 48 hours postinjury, the ARI gradually returned toward normal. During the first 12 hours postinjury, the mean value for ARI was 2.8 1.9, and in only 62 (51%) of the patients was the ARI within the normal range. At the nadir of the ARI, which occurred at 36 to 48 hours postinjury, the mean ARI was 1.7 1.1, and only 16 (13%) of the patients had an ARI in the range of normal autoregulation. From Day 2 throughout the remainder of the study, the ARI gradually increased toward normal in most patients. Values for the ARI similar to the initial values were obtained on Days 5 (ARI 2.5 1.4) through 10 (ARI 2.8 1.2). Normal values of dynamic testing of autoregulation were found in only 25 patients on Day 10 (55% of patients tested on this day). This change in the ARI over time after TBI was highly significant (time effect p 0.001, repeated-measures ANOVA), and the mean values for the ARI on Days 2 to 4 were significantly different from the values obtained during the first 12 hours postinjury (p 0.05, Tukey test). This characteristic time course of the ARI (averaged results from the left and right sides for each patient expressed as the median and the 25th and 75th percentile for the group results) is illustrated in Fig. 2. Only 20 of the 122 patients required even temporary treatment with vasopressors during the first 48 hours postinjury, and this characteristic pattern of change in the ARI over time was independent of the need for vasopressor support of the BP. The baseline MCA BFV, measured before the testing of pressure autoregulation, also changed significantly over time (time effect, p 0.001); the MCA BFV was lowest on Day 1 (52 22 cm/second), increased to 71 30 cm/second on Day 2, and peaked at 72 hours after injury (83 28 cm/second). These changes in MCA BFV over time are shown in Fig. 2 (averaged results from the left and right sides of each patient are expressed as the median and the 25th and 75th percentiles for the group results). Changes in other physiological parameters were also observed over time, but in contrast with the ARI the trends were toward gradual improvement throughout the 10 days of the study, with the most abnormal values occurring during the first 12 hours postinjury. Changes in CPP, MABP, ICP, and ETCO 2 are summarized in Table 2. The CPP and MABP gradually increased over the entire study period (time effect, p 0.001 for both parameters), but most of these changes occurred well after 48 hours postinjury. Values for CPP and MABP were nearly constant during the TABLE 2 Time course of changes in physiological variables* Parameter Overall (mm Hg) Day 1 Day 2 Day 5 Day 10 Time Effect MABP 86 16 89 11 92 13 91 12 0.001 CPP 70 18 71 13 74 16 83 13 0.001 ICP 17 8 20 5 19 8 18 4 0.26 ETCO 2 36 4 36 4 36 5 37 5 0.46 *Values are expressed as the mean SD. Statistical analysis was performed using one-way repeated-measures ANOVA. Significantly different from Day 1. first 48 hours, when the ARI continued to decrease. The mean values for ICP were slightly higher on Days 2 to 10 than on Day 1, but these differences were not significant (time effect, p = 0.26). The ETCO 2 was nearly constant throughout the study period, averaging 34 6 mm Hg during the first 12 hours and 36 5 mm Hg on Day 10 (time effect, p = 0.46). Changes During the First 48 Hours Postinjury In these studies of time course, the dynamic pressure autoregulation consistently deteriorated during the first 36 to 48 hours postinjury. This worsening of the ARI occurred in the setting of a constant MABP and CPP, and an apparent improvement in baseline CBF as estimated using the MCA BFV. To examine this phenomenon more closely, the changes in all physiological parameters that were available in this set of patients were examined. Because some of the physiological parameters, such as CBF and brain tissue PO 2, were not available in all patients, the average values for Days 1 and 2 were compared to maximize the number of values available for analysis. The mean value for the ARI decreased from 2.8 1.9 on Day 1 to 1.7 1.1 on Day 2 (p 0.001, paired t-test). The ARI on Day 2 was less than that on Day 1 in 92% of patients; the ARI was normal on Day 1 in 53 patients (43%) but remained normal in only 16 patients (13%) on Day 2. As shown in Fig. 3, these changes in the ARI were associated primarily with changes in physiological variables that indicated an increase in baseline CBF and a decrease in baseline CVR between Days 1 and 2 postinjury. The changes in the ARI could not be explained by changes in baseline CPP or ETCO 2. The BFV in the left MCA increased from 54 24 cm/ second on Day 1 to 72 31 cm/second on Day 2 (p 0.001). The BFV in the right MCA increased from 50 20 cm/second on Day 1 to 69 30 cm/second on Day 2 (p 0.001). The mean values for MABP, ICP, CPP, and ETCO 2 were not significantly different on Days 1 and 2 (Table 2). An increase in MCA BFV may not always indicate an increase in actual CBF; vasoconstriction could also increase MCA BFV with either no change or even a decrease in CBF. To confirm that these changes in MCA BFV indicated a true change in CBF, other direct and indirect measures of CBF were examined. Measurements of global CBF by using Xe-CT scans were available for some of the patients studied. Global CBF increased from 36 15 ml/100 g/min on Day 1 (54 patients) to 47 13 ml/100 g/min on Day 2 (29 patients, p 0.05; Fig. 3). In addition, measures of 1057

R. Hlatky, et al. FIG. 3. Boxplots showing comparisons of ARI, MCA BFV, Xe- CT CBF, and CVR during Days 1 and 2 postinjury. Statistical analysis was performed using paired t-tests. FIG. 4. Boxplots showing comparisons of MCA territorial CVR, ARI, and MCA BFV during Days 1 and 2 postinjury. Statistical analysis was performed using paired t-tests. The CVR values are given as mm Hg/ml/100 g/min. local cortical CBF (thermal-diffusion CBF), and indirect measures, such as SjvO 2, AVDO 2, and brain tissue PO 2, confirmed the MCA BFV findings. As summarized in Table 3, SjvO 2 (122 patients), brain tissue PO 2 (86 patients), and thermal-diffusion CBF (46 patients) significantly increased and AVDO 2 (122 patients) significantly decreased from Day 1 to Day 2. The CMRO 2 values based on Xe-CT CBF values and the corresponding AVDO 2 values were constant during the first 2 days postinjury (p = 0.81). The CVR was calculated by the formula CVR = CPP/ CBF, and estimated by the formula CVR = CPP/BFV, using MCA BFV to substitute for CBF. The CVR decreased significantly between Day 1 and Day 2 according to both methods of calculation. The CVR values based on MCA BFV decreased from 1.42 0.45 mm Hg/cm/sec on Day 1 to 1.06 0.33 mm Hg/cm/sec on Day 2. The CVR values TABLE 3 Comparison of physiological variables on Days 1 and 2 postinjury* Parameter Day 1 Day 2 p Value SjvO 2 (%) 67 8 70 9 0.05 brain tissue PO 2 (mm Hg) 27 19 36 20 0.05 rtd-cbf (ml/100 g/min) 26 17 39 21 0.05 AVDO 2 (ml/dl) 4.3 1.5 3.4 1.2 0.001 CMRO 2 (ml/100 g/min) 1.68 0.79 1.73 0.91 0.81 * Statistical analysis was performed using paired t-tests. Abbreviation: rtd-cbf = regional thermal diffusion CBF. based on global CBF decreased from 1.92 0.69 mm Hg/ ml/100 g/min on Day 1 to 1.34 0.46 mm Hg/ml/100 g/ min on Day 2. The CVR values based on cortical CBF in the MCA territories that was calculated from the Xe-CT CBF findings showed the same trend as values based on the mean global CBF. The MCA territorial CVR decreased from 2.08 1.03 mm Hg/ml/100 g/min for the left MCA and 2.46 1.31 mm Hg/ml/100 g/min for the right MCA on Day 1 to 1.47 0.61 mm Hg/ml/100 g/min for the left and 1.55 0.7 mm Hg/ml/100 g/min for the right MCA territory (both sides, p 0.01; Fig. 4). The ARI for the left and right MCA had the same trend as territorial CVR. The ARI for the left MCA decreased from 2.8 1.6 on Day 1 to 1.9 1.1 on Day 2, whereas the ARI for the right MCA changed from 2.9 1.4 to 1.8 1.2 (both sides, p 0.001). Comparison of Physiological Variables Based on Initial MCA BFV These comparisons of changes in the ARI with those in other physiological variables indicated that cerebral pressure autoregulation was dependent on the baseline CBF or cerebrovascular tone. To study this hypothesis, patients were divided into two groups based on their MCA BFV (mean of left and right sides) during the first 12 hours postinjury. Patients with an initial MCA BFV greater than the median value of 50 cm/second were included in the high BFV group (44 patients), and those with an initial MCA BFV less than 50 cm/second were considered to belong to the low BFV group (47 patients). As shown in Fig. 5, sim- 1058

Dynamic autoregulatory response after severe head injury TABLE 4 Values for physiological variables in patients with initial MCA BFV above or below median* Group Parameter Time Inter- Group (mm Hg) Group Day 1 Day 2 action Effect MABP high 85 12 89 11 0.37 0.45 low 87 15 90 10 CPP high 71 10 73 12 0.62 0.97 low 70 18 73 16 ICP high 16 5 16 5 0.42 0.33 low 15 4 18 6 ETCO 2 high 36 4 36 4 0.69 0.82 low 36 3 35 4 * Statistical analysis was performed using two-way repeated-measures ANOVA. High denotes above median and low denotes below median. FIG. 5. Graphs showing comparisons of patients with initial MCA BFV above and below the median. The group with initial MCA BFVs below the median was significantly different from the group with initial MCA BFVs above the median value (p 0.05). The means that the change in the group with an initial MCA BFV below the median is significantly different from that in the group with an initial MCA BFV above the median value (p 0.05). Statistical analysis was performed using two-way repeated-measures ANOVA. ilar differences in global CBF were observed between the two groups. The MCA BFV in the high BFV group averaged 67 14 cm/second during the first 12 hours postinjury and 74 13 cm/second on Day 2. In the low BFV group, MCA BFV was initially 38 11 cm/second and increased to 63 23 cm/second on Day 2 (time effect, p 0.001; group time interaction, p = 0.011; two-way repeated-measures ANOVA). As shown in Fig. 5, similar findings were observed with CBF values obtained using Xe-CT scanning. The changes in the ARI mirrored the MCA BFV findings in these two groups of patients. The ARI was significantly increased in the low BFV group on Day 1: 3.3 1.2 compared with 2.4 1.1 in the high BFV group. In addition, the decrease in the ARI between Days 1 and 2 was greater in the low BFV than in the high BFV group. Table 4 demonstrates no differences between the high and low BFV groups in physiological variables such as MABP, CPP, ICP, and ETCO 2, which might confound these relationships. Discussion The results of this study indicate that many patients (49% on Day 1 and 87% on Day 2) experience impaired dynamic cerebral pressure autoregulation after severe head injury. This finding is consistent with previous studies in which static testing of autoregulation was used; a similar incidence of abnormalities, ranging from 41 to 83% of patients, was reported in these studies. 4,6,17,22 Muizelaar, et al., 17 demonstrated disturbed autoregulation in 15 of 37 measurements in pediatric patients, whereas Cold and Jensen 6 observed regional loss of autoregulation, indicated by a 20% flow increase, in 29 of 35 studies. Impaired autoregulation has even been reported after mild TBI. 28 Junger, et al., 11 demonstrated the use of dynamic testing for poorly functioning or absent cerebral autoregulation in 28% of 29 patients. The ability serially to monitor and grade the autoregulatory response over time is unique to dynamic autoregulation methods, and the contribution of this study is the unexpected evolution of autoregulation that was observed over time. Although the lowest BFV and CBF values occurred during the initial 12 hours postinjury, the ability to autoregulate continued to worsen during the first 36 to 48 hours. The clinical implications of these findings are important. It is likely that most secondary insults occur in the first few hours postinjury, because this is when most hemodynamic instability occurs. 12,15 These studies indicate that the inability of the injured brain to compensate normally for hypotensive insults persists well beyond the initial 6 to 12 hours, and that vigilance should continue throughout the first few days after injury. The underlying cause of impaired autoregulation after TBI is not known. Even the mechanism of normal autoregulation is poorly understood, and is probably complex. There are four general theories about the mechanism of normal autoregulation. 1) The myogenic theory hypothesizes that vascular smooth-muscle cells have an intrinsic ability to detect and respond to varying perfusion pressures. The underlying mechanisms are thought to be an activation of stretch-sensitive channels in smooth-muscle cells that causes an intracellular calcium increase and therefore constriction. 2) The metabolic theory proposes that local metabolic factors control autoregulation. 3) The neurogenic theory hypothesizes that perivascular nerves play a role in CBF regulation. 4) The fourth theory supposes that endothelial factors may also play a role in autoregulation; in some experimental models, an intact endothelium is necessary for pressure autoregulation. 1059

R. Hlatky, et al. Trauma could potentially alter any or all of these factors. A myogenic response to intravascular pressures can be demonstrated in isolated cerebral vessels, and an analogous evolution of impairment in the myogenic response over time has been observed in one experimental TBI model. 8 The dilation of cerebral (pial) arterioles in response to increased venous pressures and the reversal of this response by local hyperoxia lend further support to the hypothesis that a metabolic component of cerebral autoregulation involves an O 2 -sensitive mechanism. 29 The other physiological parameters that were analyzed in this study revealed no confounding factors that might be responsible for the impaired autoregulatory response. A reduction in CPP could reduce the ability to autoregulate; however, CPP was well maintained in these patients with TBI and did not change significantly as the autoregulatory response deteriorated over the first 2 days postinjury. The PaCO 2 has an important effect on the autoregulatory response, with hyperventilation improving and hypoventilation impairing autoregulation. 14,20 All patients received mechanical ventilation throughout the first few days of the study, however, and ETCO 2 levels were kept constant. The possibility cannot be entirely excluded that at least some of the observed changes in autoregulation are induced by medications that are given to critically ill patients with TBI for anesthesia, sedation, analgesia, or control of ICP. Nevertheless, the same management protocol was used for all patients, and no single medication appeared to be responsible for the changes in the ARI that were observed during the first 48 hours. The only consistent association with the changes in the autoregulatory response was in measures of CBF and CVR. When CBF was low (and CVR was high) on admission, the dynamic autoregulatory response was better preserved. When CBF increased (and CVR decreased) over the first 48 hours postinjury, the autoregulatory response became more impaired. The dependence of autoregulation on baseline cerebrovascular tone may explain these findings. Some limitations of the dynamic autoregulatory response should be acknowledged. This procedure is more convenient for testing autoregulation in critically ill patients because it requires only a small and very brief perturbation in BP and because it can be easily done in the intensive care unit. This method, however, uses BFV in the large arteries supplying the brain to estimate CBF. Although BFV can be altered by both changes in vessel flow and in vessel diameter, this method has been well validated in healthy volunteers. Under the limited circumstance of a transient drop in BP, changes in BFV closely predict changes in flow measured by more standard methods. 2 Another concern with this testing procedure is that dynamic changes might not predict the ability of the brain to maintain a constant CBF over longer periods of time. Nonetheless, the dynamic ARI has been shown to correlate closely with the findings of static autoregulation testing. 28 Conclusions The ability to autoregulate pressure is commonly impaired in patients with severe TBI, and this vulnerability persists and even increases after the 1st day postinjury. Additional studies are needed to determine if impaired autoregulation leads to an increased risk for secondary ischemic insults, but this seems to be a reasonable hypothesis. If so, dynamic autoregulation testing, which can be performed serially in the intensive care unit, could be a method for targeting preventive treatment in patients who are at greatest risk, or for determining in individual patients how long treatment should be continued. References 1. Aaslid R, Lindegaard KF, Sorteberg W, et al: Cerebral autoregulation dynamics in humans. Stroke 20:45 52, 1989 2. Aaslid R, Newell DW, Stooss R, et al: Assessment of cerebral autoregulation dynamics from simultaneous arterial and venous transcranial Doppler recordings in humans. Stroke 22: 1148 1154, 1991 3. Bouma GJ, Muizelaar JP, Stringer WA, et al: Ultra-early evaluation of regional cerebral blood flow in severely head-injured patients using xenon-enhanced computerized tomography. J Neurosurg 77:360 368, 1992 4. Bruce DA, Langfitt TW, Miller JD, et al: Regional cerebral blood flow, intracranial pressure, and brain metabolism in comatose patients. J Neurosurg 38:131 144, 1973 5. Chesnut RM, Marshall LF, Klauber MR, et al: The role of secondary brain injury in determining outcome from severe head injury. J Trauma 34:216 222, 1993 6. Cold GE, Jensen FT: Cerebral autoregulation in unconscious patients with brain injury. Acta Anaesthesiol Scand 22:270 280, 1978 7. Czosnyka M, Smielewski P, Kirkpatrick P, et al: Monitoring of cerebral autoregulation in head-injured patients. Stroke 27: 1829 1834, 1996 8. Golding EM, Contant CF Jr, Robertson CS, et al: Temporal effect of severe controlled cortical impact injury in the rat on the myogenic response of the middle cerebral artery. J Neurotrauma 15: 973 984, 1998 9. Gopinath SP, Robertson CS, Contant CF: Jugular venous desaturation and outcome after head injury. J Neurol Neurosurg Psychiatry 57:717 723, 1994 10. Jennett B, Bond M: Assessment of outcome after severe brain damage. Lancet 1:480 484, 1975 11. Junger EC, Newell DW, Grant GA, et al: Cerebral autoregulation following minor head injury. J Neurosurg 86:425 432, 1997 12. Kelly DF, Martin NA, Kordestani R, et al: Cerebral blood flow as a predictor of outcome following traumatic brain injury. J Neurosurg 86:633 641, 1997 13. Kety SS, Schmidt CF: The nitrous oxide method for quantitative determination of cerebral blood flow in man: theory, procedure and normal values. J Clin Invest 27:476 483, 1948 14. Marion DW, Bouma GJ: The use of stable xenon-enhanced computed tomographic studies of cerebral blood flow to define changes in cerebral carbon dioxide vasoresponsivity caused by a severe head injury. Neurosurgery 29:869 873, 1991 15. Martin NA, Patwardhan RV, Alexander MJ, et al: Characterization of cerebral hemodynamic phases following severe head trauma: hypoperfusion, hyperemia, and vasospasm. J Neurosurg 87: 9 19, 1997 16. Miller JD, Piper IR, Dearden NM: Management of intracranial hypertension in head injury: matching treatment with cause. Acta Neurochir Suppl 57:152 159, 1993 17. Muizelaar JP, Ward JD, Marmarou A, et al: Cerebral blood flow and metabolism in severely head-injured children. Part 2: Autoregulation. J Neurosurg 71:72 76, 1989 18. Newell DW, Aaslid R, Lam A, et al: Comparison of flow and velocity during dynamic autoregulation testing in humans. Stroke 25:793 797, 1994 1060

Dynamic autoregulatory response after severe head injury 19. Newell DW, Aaslid R, Stooss R, et al: Evaluation of hemodynamic responses in head injury patients with transcranial Doppler monitoring. Acta Neurochir 139:804 817, 1997 20. Newell DW, Weber JP, Watson R, et al: Effect of transient moderate hyperventilation on dynamic cerebral autoregulation after severe head injury. Neurosurgery 39:35 44, 1996 21. 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 22. Overgaard J, Mosdal C, Tweed WA: Cerebral circulation after head injury. Part 3: Does reduced regional cerebral blood flow determine recovery of brain function after blunt head injury? J Neurosurg 55:63 74, 1981 23. Robertson CS, Contant CF, Gokaslan ZL: Cerebral blood flow, arteriovenous oxygen difference, and outcome in head injured patients. J Neurol Neurosurg Psychiatry 55:594 603, 1992 24. Robertson CS, Valadka AB, Hannay HJ, et al: Prevention of secondary ischemic insults after severe head injury. Crit Care Med 27:2086 2095, 1999 25. Strebel S, Lam AM, Matta B, et al: Dynamic and static cerebral autoregulation during isoflurane, desflurane and propofol anesthesia. Anesthesiology 83:66 76, 1995 26. Strebel S, Lam AM, Matta BF, et al: Impaired cerebral autoregulation after mild brain injury. Surg Neurol 47:128 131, 1997 27. Teasdale G, Jennett B: Assessment of coma and impaired consciousness. A practical scale. Lancet 2:81 84, 1974 28. Tiecks FP, Lam AM, Aaslid R, et al: Comparison of static and dynamic cerebral autoregulation measurements. Stroke 26: 1014 1019, 1995 29. Wei EP, Kontos HA: Increased venous pressure causes myogenic constriction of cerebral arterioles during local hyperoxia. Circ Res 55:249 252, 1984 Manuscript received June 29, 2001. Accepted in final form June 26, 2002. This work was supported by National Institutes of Health Grant No. PO1-NS38660. Address reprint requests to: Roman Hlatky, M.D., Department of Neurosurgery, Baylor College of Medicine, 6560 Fannin, Suite 944, Houston, Texas 77030. email: romanhlatky@hotmail.com. 1061