Secondary Logo

Journal Logo

Transcranial Doppler Sonography Indicates Critical Brain Perfusion During Hemorrhagic Hypotension in Dogs

Werner, Christian MD; Hoffman, William E. PhD; Kochs, Eberhard MD, MSc; Albrecht, Ronald F. MD; am Esch, Jochen Schulte MD

Neurosurgical Anesthesia

This study investigated the effects of hemorrhagic hypotension on cerebral blood flow velocity and brain electrical activity (by electroencephalogram [EEG]).Eleven mongrel dogs were anesthetized with isoflurane (1 minimum alveolar anesthetic concentration [MAC]) and catheters were placed into both femoral arteries and veins for mean arterial blood pressure (MAP) measurement, blood withdrawal, and drug administration. Brain temperature, arterial blood gases, and pH were maintained constant. EEG was recorded from temporoparietal recording sites versus a frontal reference. A pulsed transcranial Doppler (TCD) probe (2 MHz, Transpect Trademark, Medasonics) was placed on the dura via a temporal bone window to measure mean (Vmean, cm/s) and diastolic blood flow velocity (Vdiast, cm/s) in the middle cerebral artery. At the end of the surgical preparation, isoflurane was discontinued and all animals received fentanyl (bolus, 25 micro gram/kg intravenously (IV); infusion, 50 micro gram centered dot kg-1 centered dot h-1 IV) plus 50% N2 O/O2 during 30 min of equilibration. After recordings of baseline data, the dogs were hemorrhaged at a rate of 80-100 mL/min. The observation interval was 14 min. EEG spectral edge frequency (SEF 95%) and Vmean did not change when MAP was decreased from 109 +/- 10 to 63 +/- 7 mm Hg. This indicates preserved neuronal function and intact autoregulation of cerebral blood flow. Below MAP of 49 +/- 9 mm Hg, a shift of the EEG to lower frequencies was associated with decreases in Vmean and Vdiast. EEG burst suppression occurred at a MAP of 31 +/- 7 mm Hg, paralleled by a loss of the diastolic flow velocity pattern. The results of this study are consistent with findings in patients with increased intracranial pressure where the loss of diastolic blood flow velocity occurs at a level when cerebral perfusion decreases below the cerebral ischemic threshold.

(Anesth Analg 1995;81:1203-7)

University of Illinois College of Medicine, Chicago, Illinois (Hoffman, Albrecht), and University Hospital Eppendorf, Hamburg, Germany (Werner, Kochs, Shulte).

Section Editor: Donald S. Prough.

Accepted for publication June 28, 1995.

Address correspondence and reprint requests to Christian Werner, MD, Institute of Anesthesiology, Technische Universitat Munchen, Klinikum rechts der Isar, Ismaninger Strasse 22, 81675 Munchen, Germany.

Monitoring of vital physiologic functions during anesthesia and critical care may reduce perioperative morbidity and mortality [1,2]. Central nervous system monitoring is performed specifically to observe physiologic alterations associated with surgery, trauma, or secondary insults. Monitoring of cerebral ischemia is of particular importance during anesthesia or critical care since clinical evaluation of adequate cerebral blood flow (CBF) is usually impossible in anesthetized, sedated, or comatose patients. Measurements of the electroencephalogram (EEG) and intracranial pressure (ICP) have improved cerebral monitoring but these techniques permit only indirect conclusions concerning cerebral ischemia because many other factors may interfere with the ability of EEG or ICP to assess cerebral perfusion [2].

Transcranial Doppler sonography (TCD) is a noninvasive and real-time measurement technique of blood flow velocity in major basal cerebral arteries. Studies in patients with focal [3] and incomplete or complete global cerebral ischemia [4-6] suggest that occlusion of cerebral arteries or increases in ICP produce concurrent decreases in the TCD blood flow velocity. However, a specific TCD flow velocity pattern indicating cerebral ischemia has not yet been determined. The purpose of the present study was to evaluate the effects of hemorrhagic hypotension on the EEG in relation to TCD blood flow velocity and to identify TCD flow patterns associated with EEG burst suppression (as an end-point indicating cerebral ischemia).

Back to Top | Article Outline


These experiments were approved by the Institutional Animal Care Committee of the University of Illinois at Chicago. Eleven mongrel dogs (27-34 kg) were anesthetized with 1.4% isoflurane end-tidal and 50% N2 O in O2, intubated, and mechanically ventilated. Vecuronium was used for neuromuscular block. Heart rate (bpm) was recorded by electrocardiogram. Catheters were inserted into both femoral arteries and veins for continuous measurement of mean arterial (MAP, mm Hg) and central venous blood pressure (mm Hg), arterial blood sampling, hemorrhage, and drug administration. The left zygomatic arch was resected and a cranial window was drilled in the temporal bone to visualize the middle cerebral artery (MCA). A pulsed 2-MHz Doppler ultrasound probe was placed on the intact dura to measure blood flow velocity in the proximal segment of the MCA (insonation depth, 20-24 mm). The probe was then fixed in a frame to keep the depth and angle of insonation constant. A catheter was inserted into the lateral cerebral ventricle for measurement of ICP (mm Hg). Cerebral perfusion pressure (CPP, mm Hg) was calculated as CPP = MAP--ICP. Brain temperature was measured using a thermistor probe (Yellow Springs Instrument Co., Yellow Springs, OH) between the dura and the skull at the site of the cranial window. Brain temperature was maintained constant by servocontrol using a heat lamp placed over the head of the animals and connected to the thermistor. Arterial blood gases and pHa were maintained constant by adjusting ventilation and infusing sodium bicarbonate. At the end of surgery, all incisions were infiltrated with 0.25% bupivacaine and the administration of isoflurane was discontinued. During an equilibration period of 30 min, the isoflurane was completely eliminated and anesthesia was continued using fentanyl (bolus, 25 micro gram/kg intravenously; infusion, 50 micro gram centered dot kg-1 centered dot h-1 intravenously) plus 50% N2 O/O2 in all animals. After recordings of baseline data, the stopcock of one femoral artery was opened and the dogs hemorrhaged at a rate of 80-100 mL/min.

The EEG was recorded from both hemispheres at parietotemporal versus frontal recording sites using screw electrodes (Biosignal-Verstarker Trademark; nbnelectronics, Germany). The electrooculogram was recorded from supraorbital versus infraorbital leads for artifact control. Inter-electrode impedances were less than 5 kOhm. Bandpass for EEG and electrooculogram recordings was set at 0.5-45 Hz. After digitization (100/s; ced 1401 Trademark; Cambridge Electronics, England) and fast Fourier transformation (epoch length 5.2 s), EEG power spectra were analyzed. Spectral edge frequency (SEF) was calculated as the frequency where 95% of the EEG power is at lower frequencies. The occurrence of burst suppression EEG with isoelectric periods of more than 3 s was considered the cerebral ischemic threshold [7].

MCA blood flow velocity was measured continuously using a pulsed 2-MHz transcranial Doppler system (Transpect Trademark; Medasonics, Mountain View, CA). The system operates with ultrasonic intensities up to 96 mW/cm2 and pulse repetition frequencies between 4.96 kHz and 20.52 kHz. A range-gate is used to adjust the ultrasonic focus electronically in steps of 2.0 mm. The axial extension of the sample volume is 7.5 mm according to a burst width of 13 mu s. A high-pass filter of 150 Hz is set for signal registration. After Doppler shift calculation and flow direction discrimination, signals are computed using spectral analysis by 256-point fast Fourier transformation, averaging cycles of 4-20 s. The flow velocity profile is displayed in real-time on a video monitor. The instantaneous mean (Vmean, cm/s) and diastolic blood flow velocity (Vdiast, cm/s) are digitally displayed for each flow spectrum Figure 1.

Figure 1

Figure 1

All data are reported as mean +/- SD. All physiologic variables were compared using repeated-measures analysis of variance. Post hoc comparisons of physiologic variables across time were made using Tukey's test, corrected for multiple comparisons. Statistical significance was assumed at P < 0.05.

Back to Top | Article Outline


Arterial blood gases and arterial pHa did not change over time with the exception of decreases in PaCO2 and pH during hemorrhagic hypotension at the end of the experiment Table 1. MAP decreased significantly over time according to the protocol, and heart rate increased with hemorrhage. Central venous pressure and CPP were decreased 8 min and 6 min after start of hemorrhage. ICP was decreased at the end of the experiment. Figure 2 shows changes in SEF and MCA blood flow velocity as a function of hemorrhagic hypotension. SEF was constant during the decline of MAP to a level of 56 +/- 4 mm Hg with a shift to lower frequencies (3-6 Hz) between 56 +/- 4 mm Hg and 49 +/- 9 mm Hg. SEF was reduced compared to baseline at MAP of less than 49 +/- 9 mm Hg and EEG burst suppression occurred at MAP of 31 +/- 7 mm Hg. Vmean and Vdiast were constant during the decline of MAP to levels of 63 +/- 7 mm Hg and 71 +/- 9 mm Hg, respectively. Below these MAP levels, TCD blood flow velocity was progressively decreased (P < 0.05). In each animal the occurrence of EEG burst suppression was associated with loss of the diastolic flow pattern.

Table 1

Table 1

Figure 2

Figure 2

Back to Top | Article Outline


The present results show a close relation between changes in SEF and TCD blood flow velocity during acute hemorrhagic hypotension. SEF and TCD velocity were constant during the decrease of MAP to a level of 55% from baseline. This is consistent with preserved CBF autoregulation [6,8]. Further decreases in MAP produced a shift of the EEG to lower frequencies parallel to decreases in Vmean and Vdiast. At MAP levels of 31 +/- 7 mm Hg, EEG burst suppression occurred. Concurrently, a loss of the diastolic flow velocity pattern was seen. This shows that the diastolic TCD blood flow velocity and neuronal function are closely correlated. Although CBF was not measured during the present study these results suggest that the diastolic flow pattern is an indicator of severe cerebral ischemia as a function of CPP.

Previous studies have shown that arterial hypotension beyond the lower limit of CBF autoregulation produces typical changes in the EEG pattern that are related to progressive cerebral ischemia and concurrent neuronal dysfunction. Stockard et al. [9] studied the relation between different arterial blood pressure levels during cardiopulmonary bypass and postoperative neurologic status. They found that EEG slowing or isoelectricity and postoperative neurologic deficits were correlated in a time-dependent fashion. This is consistent with an investigation in patients undergoing carotid endarterectomy where 50% decreases in SEF for a period of 10 min were correlated with the development of postoperative neurologic deficits [7]. Based on these investigations, the SEF was used here to measure the effects of progressive systemic hypotension on neuronal function in comparison to cerebral hemodynamics assessed by the TCD technique although the time frame of the present experiment may differ from previous studies.

The induction of ischemic EEG changes may occur at different levels of MAP. Kovach and Sandor [10] found EEG impairment at MAP levels of 50-60 mm Hg with subsequent isoelectricity at MAP levels of 30-40 mm Hg in unanesthetized cats and dogs. In contrast, Gregory et al. [11] have shown that brain electrical activity did not change over a MAP range of 120-40 mm Hg when chloralose-anesthetized cats were subjected to graded hypotension. Below this pressure the EEG became slower and isoelectricity occurred within a MAP range of 15-30 mm Hg. This is consistent with the present data showing EEG slowing at MAP of less than 49 +/- 9 mm Hg and EEG burst suppression at MAP of 31 +/- 7 mm Hg. Yashon et al. [12] have shown that neuronal activity may be preserved in barbiturate-anesthetized dogs despite rapid reduction of MAP as low as 30 mm Hg. The differences in the cerebral ischemic threshold seen in these studies may be a function of rapid versus graded induction of systemic hypotension with recruitment of collateral perfusion with graded hypotension. Cerebral metabolic suppression using barbiturates and reduction of sympathetic tone with the use of ganglionic blockade may additionally increase the neuronal tolerance during hypotensive challenges [12,13].

Ischemic EEG changes occur at CBF levels of less than 22 mL centered dot 100 g-1 centered dot min-1[11,14]. This threshold reflects a reduction of CBF by 50%-60%. CBF levels of less than 15 mL centered dot 100 g-1 centered dot min-1 induce isoelectricity with irreversible neuronal injury occurring at CBF levels of less than 6 mL centered dot 100 g-1 centered dot min-1[11,14]. MacKenzie et al. [15] found a high variability in the relation between CBF and arterial hypotension. MAP levels of 35 mm Hg were related to an average CBF of 28 mL centered dot 100 g-1 centered dot min-1 which would be above the ischemic threshold. However, 50% of the cortical territories developed local/regional cerebral circulatory arrest within the MAP range of 59 mm Hg to 18 mm Hg. This indicates that regional cerebral ischemia may occur within a wide range of MAP and that the individual ischemic threshold cannot be identified by MAP monitoring alone.

Several studies have evaluated the changes of the TCD blood flow velocity pattern during various ischemic challenges. Clinical investigations by Padayachee et al. [16] and Naylor et al. [17] suggest a linear correlation between carotid stump pressure and decreases in Vmean during internal carotid cross-clamping. This supports results by Jorgensen et al. [3] who found stump pressure values of <40 mm Hg together with Vmean of less than 30 cm/s as an accurate indicator of cross-clamp CBF values of less than 20 mL centered dot 100 g-1 centered dot min-1. Halsey et al. [18] compared TCD, EEG, and CBF measurements in patients undergoing carotid endarterectomy. Their results suggest a threshold of <15 cm/s of the MCA mean blood flow velocity to indicate focal cerebral ischemia produced by internal carotid cross-clamping. However, this threshold for blood flow velocity was not always paralleled by either low CBF or ischemic EEG changes, indicating a low specificity and sensitivity. Giller et al. [19] and Spencer et al. [20] found a close correlation between decreases in average MCA blood flow velocity (>65%) and changes in CBF, stump pressure as well as clinical symptoms of cerebral ischemia in awake patients undergoing balloon occlusion tests of the internal carotid artery. Although these studies show that monitoring of relative or absolute changes in mean blood flow velocity may be inadequate to reproducibly detect focal or hemispheric ischemia, the present experiments show a close relation between changes in SEF and the diastolic blood flow velocity during incomplete global ischemia.

With progressive hemorrhagic hypotension, the induction of EEG burst suppression was paralleled by decreases of the diastolic MCA blood flow velocity to zero. This is consistent with TCD flow patterns observed in laboratory animals and patients with head injury or after subarachnoid hemorrhage, where decreases in Vdiast and a reduction of the cerebral arteriographic filling occurred parallel to increases in ICP [5,6,21-25]. Loss of the diastolic flow velocity pattern may be due to collapse of downstream vessels as diastolic arterial blood pressure decreases below the critical closing pressure [26]. It is also possible that diastolic circulatory arrest is a function of CPP reduction along the still patent vascular bed [27]. Studies in patients have shown that transient cessation of diastolic TCD flow may occur in the early stages of subarachnoid hemorrhage [28], cerebral anoxia [29], or with acute intracranial hematoma [30]. This early diastolic cerebral circulatory arrest is potentially reversible. The interval between loss of diastolic flow and irreversible neuronal injury may last as long as 150 min [30]. This suggests that loss of the diastolic flow pattern represents a level of CBF at which neuronal function is deteriorated while structural neuronal metabolism may be still maintained. However, it is possible that a combination of hypoxemia and hypoperfusion may lead to EEG burst suppression before loss of diastolic blood flow velocity occurs (i.e., TCD blood flow velocity is nonquantitative with respect to defining outcome).

In conclusion, the present results show that SEF and TCD blood flow velocity remain constant during acute hemorrhagic hypotension to a level of 55% from baseline. This indicates maintained cerebral autoregulation with preserved neuronal function. MAP levels sufficient to produce EEG burst suppression were paralleled by loss of the diastolic flow velocity pattern. This shows that the diastolic TCD blood flow velocity and neuronal function are closely correlated.

The authors wish to thank Richard Ripper and George Dominguez for their excellent technical assistance.

Back to Top | Article Outline


1. Eichhorn JH. Prevention of intraoperative anesthesia accidents and related severe injury through safety monitoring. Anesthesiology 1989;70:572-7.
2. Jones PA, Andrews PJD, Midgley S, et al. Measuring the burden of secondary insults in head-injured patients during intensive care. J Neurosurg Anesth 1994;6:4-14.
3. Jorgensen LG, Schroeder TV. Transcranial Doppler for detection of cerebral ischemia during carotid endarterectomy. Eur J Vasc Surg 1992;6:142-7.
4. Hassler W, Steinmetz H, Galowski J. Transcranial Doppler ultrasonography in raised intracranial pressure and in intracranial circulatory arrest. J Neurosurg 1988;68:745-51.
5. Hassler W, Steinmetz H, Pirschel J. Transcranial Doppler study of intracranial circulatory arrest. J Neurosurg 1988;71:195-201.
6. Newell DW, Aaslid R, Lam A, et al. Comparison of flow and velocity during dynamic autoregulation testing in humans. Stroke 1994;25:793-7.
7. Rampil JI, Holzer JA, Quest DO, et al. Prognostic value of computerized EEG analysis during carotid endarterectomy. Anesth Analg 1983;62:186-92.
8. Paulson OB, Strandgaard S, Edvinsson L. Cerebral autoregulation. Cerebrovasc Brain Metab Rev 1990;2:161-91.
9. Stockard JJ, Bickford RG, Schauble JF. Pressure-dependent cerebral ischemia during cardiopulmonary bypass. Neurology 1973;23:521-9.
10. Kovach AGB, Sandor P. Cerebral blood flow and brain function during hypotension and shock. Annu Rev Physiol 1976;38:571-96.
11. Gregory PC, McGeorge AP, Fitch W, et al. Effects of hemorrhagic hypotension on the cerebral circulation. II. Electrocortical function. Stroke 1979;10:719-23.
12. Yashon D, Locke GE, Bingham WG Jr, et al. Cerebral function during profound oligemic hypotension in the dog. J Neurosurg 1971;34:494-9.
13. Werner C, Hoffman WE, Thomas C, et al. Ganglionic blockade improves neurologic outcome from incomplete ischemia in rats: partial reversal by exogenous catecholamines. Anesthesiology 1990;73:923-9.
14. Branston NM, Symon L, Crockard HA, Pasztor E. Relationship between the cortical evoked potential and local cortical blood flow following acute middle cerebral artery occlusion in the baboon. Exp Neurol 1974;45:195-208.
15. MacKenzie ET, Farrar JK, Fitch W, et al. Effects of hemorrhagic hypotension on the cerebral circulation I. Cerebral blood flow and pial arteriolar caliber. Stroke 1979;10:711-8.
16. Padayachee TS, Gosling RG, Bishop CC, et al. Monitoring middle cerebral artery blood flow velocity during carotid endarterectomy. Br J Surg 1986;73:98-100.
17. Naylor AR, Wildsmith JAW, McClure J, et al. Transcranial Doppler monitoring during carotid endarterectomy. Br J Surg 1991;78:1264-8.
18. Halsey JH, McDowell HA, Gelmon S, Morawetz RB. Blood velocity in the middle cerebral artery and regional cerebral blood flow during carotid endarterectomy. Stroke 1989;20:53-8.
19. Giller CA, Mathews D, Walker B, et al. Prediction of tolerance to carotid artery occlusion using transcranial Doppler ultrasound. J Neurosurg 1994;81:15-9.
20. Spencer MP, Thomas IG, Moehring MA. Relation between middle cerebral artery blood flow velocity and stump pressure during carotid endarterectomy. Stroke 1992;23:1439-45.
21. Klingelhofer J, Conrad B, Benecke R, et al. Evaluation of intracranial pressure from transcranial Doppler studies in cerebral disease. J Neurol 1988;235:159-62.
22. Klingelhofer J, Sander D, Holzgraefe M, et al. Cerebral vasospasm evaluated by transcranial Doppler ultrasonography at different intracranial pressures. J Neurosurg 1991;75:752-8.
23. De Bray JM, Saumet JL, Berson M, et al. Acute intracranial hypertension and basilar artery blood flow velocity recorded by transcranial Doppler sonography: an experimental study in rabbits. Clin Physiol 1992;12:19-27.
24. Barzo P, Doczi T, Csete K, et al. Measurements of regional cerebral blood flow and blood flow velocity in experimental intracranial hypertension: infusion via the cisterna magna in rabbits. Neurosurgery 1991;28:821-5.
25. Chan K-H, Miller JD, Dearden NM, et al. The effect of changes in cerebral perfusion pressure upon middle cerebral artery blood flow velocity and jugular bulb venous oxygen saturation after severe brain injury. J Neurosurg 1992;77:55-61.
26. Dewey RC, Pieper HP, Hunt WE. Experimental cerebral hemodynamics. Vasomotor tone, critical closing pressure, and vascular bed resistance. J Neurosurg 1974;41:597-606.
27. Auer LM, Ishiyama N, Hodde KC. Effect of intracranial pressure on bridging veins in rats. J Neurosurg 1987;67:263-8.
28. Grote E, Hassler W. The critical first minutes after subarachnoid hemorrhage. Neurosurgery 1988;22:654-61.
29. Kirkham FJ, Levin SD, Padayachee TS, et al. Transcranial pulsed Doppler ultrasound findings in brain stem death. J Neurol Neurosurg Psychiatry 1987;50:1504-13.
30. Shiogai T, Sato E, Tokitsu M, et al. Transcranial Doppler monitoring in severe brain damage: relationships between intracranial haemodynamics, brain dysfunction and outcome. Neurol Res 1990;12:205-13.
© 1995 International Anesthesia Research Society