It is not clear whether the increase of cerebral blood flow (CBF) produced by volatile anesthetics is maintained during prolonged anesthesia. In a previous study, the authors found that CBF equivalent, an index of flow-metabolism relationship, was stable over 3 h, suggesting no decay over time in CBF for 3 h during volatile anesthesia in humans. However, it may be possible that CBF changes in a parallel fashion to functional metabolic changes. In this study, to estimate the response of CBF to three volatile anesthetics, the authors used transcranial Doppler (TCD) ultrasonography to measure time-averaged mean velocity in the middle cerebral artery (Vmca).


Twenty-four surgical patients were randomly assigned to three groups to receive halothane, isoflurane, or sevoflurane (eight patients, each). End-tidal concentration of the selected volatile anesthetic was maintained at 0.5, 1.0, and 1.5 MAC before surgery and then at 1.5 MAC during surgery, which lasted more than 3 h. Normothermia and normocapnia were maintained. Mean arterial blood pressure was kept above 70 mmHg, using phenylephrine infusion, if necessary. TCD recordings of the Vmca were performed continuously.


Vmca at 0.5 MAC of halothane, isoflurane, and sevoflurane was 49 +/- 19, 57 +/- 8, and 48 +/- 13 cm/s, respectively. Halothane significantly (P < 0.01) increased Vmca in a dose-dependent manner (0.5, 1.0, 1.5 MAC), whereas isoflurane and sevoflurane produced no significant dose-related changes. At 1.5 MAC for 3 h, Vmca changed significantly (P < 0.05) for the time trends, but it did not exhibit decay over time with all drugs. During burst suppression, observed electroencephalographically (EEG) on patients during isoflurane and sevoflurane anesthesia, the onset of a burst increased Vmca (approximately 5-30 cm/s), which was maintained for the duration of the burst.


The results indicate that there was no decay in Vmca over time during prolonged (3 h) inhalation of volatile anesthetics at 1.5 MAC in humans. The fluctuation of Vmca during burst suppression on EEG at 1.5 MAC indicates that the flow-metabolism coupling occurred.

Key words: Anesthetics, volatile: halothane, isoflurane; sevoflurane, Brain: blood flow velocity, Electroencephalograph: burst suppression, Surgical stimulation: anesthetic interaction with time.

In general, most volatile anesthetics are considered to be cerebral vasodilators, although it is controversial whether the increase, if any, of cerebral blood flow (CBF) by volatile anesthetics is maintained for a prolonged period of anesthesia. In animals, cerebral hyperemia induced by isoflurane and halothane spontaneously decreases over time. [1-6]Some reports have failed to observe the gradual decrease in CBF over time. [7-10]In the previous human study, we found that the elevated CBF equivalent, an index of flow-metabolism relationship, was preserved during prolonged anesthesia with volatile anesthetics, [11]while surgery is ongoing. However, one may have to consider that the CBF equivalent does not provide real CBF values.

Transcranial Doppler (TCD) ultrasonography provides a rapid, noninvasive, and continuous assessment of blood flow velocity of basal cerebral artery of interest. Although flow velocity is not equivalent to CBF, changes in flow velocity reflect corresponding relative changes in CBF [12,13]and there is excellent validation. Recent human studies suggested that flow velocity in the middle cerebral artery (MCA) can be affected by a functional fluctuation induced by anesthetic itself [14]and also can be affected by arousal response (increase of cerebral metabolism) induced by nociceptive stimulation, [15]even if other physiologic variables are maintained within the normal range. To estimate the response of flow velocity to three different volatile anesthetics used, we measure time-averaged mean velocity in the MCA (V sub mca) during induction, prolonged surgery, and emergence of anesthesia.

The study protocol was approved by the Ethical Committee for Human Study of the Yamaguchi Rosai Hospital, and informed consent was obtained from each patient. Twenty-four (American Society of Anesthesiologist' (ASA) physical status I or II patients (15 men, 9 women) were randomly assigned to three groups (8 for each) receiving either halothane, isoflurane, or sevoflurane during elective surgery. Ages of the patients ranged from 20 to 73 yr. Surgical procedures included orthopedic and abdominal surgery, which lasted more than 3 h. Atropine sulfate, 0.5 mg, and midazolam, 3-5 mg, were given intramuscularly 30 min before induction. Anesthesia was induced with the selected volatile anesthetic in an air-oxygen mixture adjusted to obtain FiO2of 0.35, and inspired concentration of volatile anesthetic was increased to 3-4% over 3-4 min. Tracheal intubation was facilitated with intravenous administration of 8-mg vecuronium bromide. After intubation, end-tidal concentration of the selected volatile anesthetic was adjusted to an age-appropriate level of 0.5 MAC, [16-19]then increased to 1.0 MAC, 1.5 MAC before surgery, and maintained at 1.5 MAC for the period of surgical procedure. After the end of surgery, the concentration of volatile anesthetics was stepwise decreased (1.5, 1.0, and 0.5 MAC).

Patients' lungs were mechanically ventilated to maintain normocapnia, and FiO2was kept at 0.35. The end-tidal concentrations of carbon dioxide (ETCO2) and volatile anesthetic were continuously monitored with a calibrated infrared gas analyzer (Capnomac Ultima, Datex, Helsinki, Finland). Apart from vecuronium, no other drugs were administered during the surgical procedure except phenylephrine, which was used to maintain mean arterial blood pressure (MABP, see below). The nasopharyngeal temperature was monitored by a calibrated thermistor probe and was kept at 35.5-37.0 [degree sign] Celsius using a cooling-warming water mattress. Bilateral unipolar (earlobe as a reference electrode), frontal, and parietal electroencephalographs (EEGs) were monitored and recorded continuously (Neuropack 8, Nihon Kohden, Tokyo, Japan). The electrocardiograph (ECG) also was monitored. A 22-gauge Teflon indwelling catheter was placed in the radial artery for blood sampling and pressure measurement. The arterial pressure was measured by strain gauge transducers with the zero point at the mastoid process and was recorded on a polygraph (Lifescope 14, Nihon Kohden, Tokyo). The MABP was maintained above 70 mmHg with a continuous infusion of phenylephrine (0.1-1.0 micro gram [center dot] kg sup -1 [center dot] min sup -1), if necessary. Arterial blood samples were obtained every 60 min and analyzed for oxygen tension (PO2) and carbon dioxide tension (PCO2) with a blood gas analyzer (ABL505, Radiometer, Copenhagen, Denmark) at 37.0 [degree sign] Celsius. Hemoglobin concentration was measured spectrophotometrically (OSM3, Radiometer).

Doppler Measurements

A 2-MHz pulsed Doppler ultrasound device (Transscan, EME, Uberlingen, Germany) was used for transcranial measurements of blood flow velocity of the right MCA. TCD ultrasound signals were identified at a depth of 42-60 mm. Meticulous care was taken to ensure a constant position of the ultrasound probe by use of a suitable holder attached to the patient's head (IMP-2 monitoring probe holder, EME). Measurement of flow velocity was started shortly after induction of anesthesia. Flow velocity and blood pressure were displayed simultaneously on a video screen and continuously recorded on a microcomputer by use of the long-term monitoring option of the Doppler device. Vmcawas derived from on-line integration of the recordings of envelope curves of maximal intravascular velocity. During induction or emergence of anesthesia, equilibration time at each end-tidal concentration is 5-10 min. During operation at 1.5 MAC, Vmcarepresents an average value of those recorded for approximately 1 min at each time point (every 15 min). When EEG shows burst suppression pattern, we calculated averaged value from both Vmca, values during burst period and suppression period.

Data Analysis

Data are expressed as mean +/- SD. Between-group comparisons of demographic data were made by one-way analysis of variance (ANOVA). Gender distribution between groups was compared by chi-square analysis. For the statistical analysis on physiologic variables and Vmca, the data were separated into three parts: the initial dose-response data (induction period, 0.5, 1.0, 1.5 MAC), the time course data at 1.5 MAC, and the final dose-response data (emergence period, 1.5, 1.0, 0.5 MAC). Two-way ANOVA for repeated measures was applied on each three part. Bonferroni's post hoc test was applied for between-group comparisons as indicated. Statistical significance was assumed when P < 0.05.

A summary of demographic data is shown in Table 1. There were no significant differences among three groups. Figure 1showed MABP and heart rate during the study. There were no significant differences among three groups except MABP during induction: MABP in the sevoflurane group was significantly (P < 0.05) lower than those in the halothane and isoflurane groups during induction. The total number of patients given phenylephrine was eight, three, and four in the halothane, isoflurane, and sevoflurane groups, respectively. Hemoglobin concentration (9.4-14.6 g/dl), PaO2(99-250 mmHg), PaCO2(37-45 mmHg), ETCO2(35-48 mmHg), and nasopharyngeal temperature (35.7-37.5 [degree sign] Celsius) were stable in each patient. No statistical significant differences were observed in these data among three groups.

(Figure 2) shows the time course of the changes of Vmca. Vmcaat 0.5 MAC of halothane, isoflurane, and sevoflurane was 49 +/- 19, 57 +/- 8, and 48 +/- 13 cm/s, respectively, with no significant difference among the anesthetics. Halothane significantly (P < 0.01) increased Vmcain a dose-dependent manner (0.5, 1.0, 1.5 MAC) during induction, whereas other two anesthetics produced no dose-related increase. During 3 h anesthesia period at 1.5 MAC, ANOVA for repeated measures showed significance (P < 0.05) for the time trends for Vmcadata in all groups, but there was no decay in Vmcaover time (Figure 2). No significant differences of Vmcawere obtained among three groups at 1.5 MAC during operation. During emergence of anesthesia, Vmcasignificantly (P < 0.01) decreased in a dose-dependent manner (1.5, 1.0, 0.5 MAC) in the halothane group, whereas no significant dose-related decrease was observed in the isoflurane and sevoflurane groups.

At 1.5 MAC, EEG recordings showed frequent burst (sharp or spike wave, 50-200 micro V) and suppression in the isoflurane group, whereas in the halothane and sevoflurane groups, 12-15 Hz activities (50-90 micro V) and 9-13 Hz activities (70-100 micro V) were predominant, respectively. In the sevoflurane group, burst suppression was observed in a short time in 2 patients.

(Figure 3) depicts representative recordings of MABP, Vmca, and EEG in a patient anesthetized with 1.5 MAC isoflurane. During burst and suppression in EEG, the onset of a burst resulted in a substantial increase (approximately 5-30 cm/s) in Vmca, which was maintained for the duration of the burst in all patients.

This study demonstrated in humans that Vmcasignificantly (P < 0.01) increased during induction or decreased during emergence in a dose-dependent manner in halothane anesthesia, whereas isoflurane and sevoflurane produced no significant dose-dependent change. Changes of Vmcaduring 3-h period at 1.5 MAC were significant (P < 0.05) for the time trends for all anesthetics, but there was no decay in Vmcaover time. It was also shown that Vmca fluctuated during burst suppression in EEG at 1.5 MAC of isoflurane (and sevoflurane), Vmcabeing increased during burst activity and decreased during suppression.

Dose-related increase in Vmcaobserved in the halothane group is in good agreement with that reported by Thiel et al. [20]With isoflurane, Bisonnette [21]showed no difference of Vmcabetween 0.5 and 1.5 MAC, and Thiel [20]showed no difference between 2 MAC and awake values. There have been many studies describing that the increase in CBF in isoflurane anesthesia, if present at all, is smaller than the changes observed with halothane. [22]Although the studies for sevoflurane are limited in number, either decreased [23,24]or unchanged [25,26]CBF was reported in animals. Low CBF values were reported in patients with ischemic cerebrovascular disease who were anesthetized with 1.5% sevoflurane and 33% nitrous oxide. [27]Recent study in healthy patients showed that Vmcawas decreased under 1.2 MAC sevoflurane anesthesia when compared with the awake value. [28]Although we have no awake value of Vmca, considering these previously mentioned results, no dose-related increase in the current study may be the case for isoflurane and sevoflurane.

Many animal studies using dogs and goats revealed that cerebral hyperemia induced by isoflurane and halothane spontaneously decreased over time. [1-6]However, some reports [7,8]have failed to observe the gradual decrease in CBF over time. There is no clear explanation for the different results reported in animals, but it may be a result of the differences of species or the technique of CBF measurement. It has been our concern whether time-dependent decay of CBF takes place in humans. There has been limited human investigation on this matter, and reported results suggest that CBF does not decrease over time during anesthesia. [9,10]However, this lack of time decay was based on two determinations in the beginning and at the end of the study only (in 1- and 2-h, [9]and 3- and 6-h [10]periods after induction). In our previous study, [11]we found that the elevated CBF equivalent, calculated every 20 min, was preserved during prolonged anesthesia (3 h) with volatile anesthetics and suggested that the increase of CBF was maintained over time. However, this indirect method has limitations: real CBF cannot be obtained, and only the global ratio of CBF-metabolism for certain time period is obtained. In the present study, we incorporated TCD technique to measure Vmcacontinuously during prolonged anesthesia. Flow velocity is not equivalent to CBF, but it is well accepted that the changes in flow velocity reflect corresponding relative changes in CBF. [12,13]Because the flow velocity with isoflurane and sevoflurane did not increase in a dose-dependent manner, it is difficult to state that an "increase" in flow velocity is maintained during isoflurane and sevoflurane anesthesia. However, no decay in Vmcaover time in all groups in this study suggests that the gradual decrease in CBF over time do not take place in humans.

The statistical significance of the changes in Vmcafor the time trends during 3-h period at 1.5 MAC was interpreted as a reflection of the fluctuation of Vmca, possibly in association with functional changes. The changes of EEG activities reflect the functional changes of the brain and may be associated with the change of CBF. During burst suppression pattern in EEG observed at 1.5 MAC isoflurane (and sevoflurane) anesthesia, the onset of a burst activity was associated with an increase of Vmca, which was maintained for the duration of the burst, and Vmcadecreased during the suppression period (Figure 3). This observation coincides with that by Lam et al. [14]The fluctuation of Vmcaat 1.5 MAC may partly be ascribed to the nociceptive stimulation. The change of flow velocity induced by surgical stimulus has been recently reported by von Knobelsdorff et al., [15]showing that flow velocity significantly increases, lasting for 3 or 11 min after skin incision in 1 and 2 MAC isoflurane-anesthetized patients. These changes were possibly a result of the increased cerebral metabolism (i.e., arousal response) but not to the increase of MABP. [15]Bisonnette et al. [21]showed that Vmcaremained stable during 1.0 MAC isoflurane anesthesia, with concomitant use of epidural local anesthetic. The fluctuation of Vmcaat 1.5 MAC may also partly be ascribed to changes in CBF associated with MABP changes because of the possibility of impaired autoregulation. [29]Taken together, the changes of Vmcaobserved appear to be the net results of the vascular and metabolic action of anesthetics and the effect of nociceptive stimulation. Our previous observation [11]that CBF equivalent was preserved is in agreement with the present one in view of existence of flow-metabolism coupling during 1.5 MAC volatile anesthesia.

In summary, the current study shows that no decay in CBF over time occurs in the patients during surgery under prolonged (3 h) inhalation of volatile anesthetics (1.5 MAC). The fluctuation of CBF observed during burst suppression in EEG at 1.5 MAC during isoflurane and sevoflurane anesthesia indicates the existence of flow-metabolism coupling.

Raichle ME, Posner JB, Plum F: Cerebral blood flow during and after hyperventilation. Arch Neurol 1970; 23:394-403.
Albrecht RF, Miletich DJ, Madala LR: Normalization of cerebral blood flow during prolonged halothane anesthesia. Anesthesiology 1983; 58:26-31.
Boarini DJ, Kassell NF, Coester HC, Butler M, Sokoll MD: Comparison of systemic and cerebrovascular effects of isoflurane and halothane. Neurosurgery 1984; 15:400-9.
Turner DM, Kassel NF, Sasaki T, Comair YG, Boarini DJ, Beck DO: Time-dependent changes in cerebral and cardiovascular parameters in isoflurane-nitrous oxide-anesthetized dogs. Neurosurgery 1984; 14:135-41.
Warner DS, Boarini DJ, Kassel NF: Cerebrovascular adaptation to prolonged halothane anesthesia is not related to cerebrospinal fluid pH. Anesthesiology 1985; 63:243-8.
Brian JE, Traystman RJ, McPherson RW: Changes in cerebral blood flow over time during isoflurane anesthesia in dogs. J Neurosurg Anesth 1990; 2:122-30.
Roald OK, Forsman M, Steen PA: The effects of prolonged isoflurane anesthesia on cerebral blood flow and metabolism in the dog. Acta Anaesthesiol Scand 1989; 33:210-3.
McPherson RW, Kirsch JR, Tobin JR, Ghaly RF, Traystman RJ: Cerebral blood flow in primates is increased by isoflurane over time and is decreased by nitric oxide synthase inhibition. Anesthesiology 1994; 80:1320-7.
Madsen JB, Cold GE, Hansen ES, Bardrum B: The effect of isoflurane on cerebral blood flow and metabolism in human during craniotomy for small supratentorial cerebral tumors. Anesthesiology 1987; 66:332-6.
Ornstein E, Young WL, Fleischer LH, Ostapkovich N: Desflurane and isoflurane have similar effects on cerebral blood flow in patients with intracranial mass lesions. Anesthesiology 1993; 79:498-502.
Kuroda Y, Murakami M, Tsuruta J, Murakawa T, Sakabe T: Preservation of the ratio of cerebral blood flow/metabolic rate for oxygen during prolonged anesthesia with isoflurane, sevoflurane, and halothane in humans. Anesthesiology 1996; 84:555-61.
Bishop CCR, Powell S, Rutt D, Browse NL: Transcranial Doppler measurement of middle cerebral artery blood flow velocity: A validation study. Stroke 1986; 17:913-5.
Kochs E, Hoffman WE, Werner C, Albrecht RF, Schulte am Esch J: Cerebral blood flow velocity in relation to cerebral blood flow, cerebral metabolic rate for oxygen, and electroencephalogram analysis during isoflurane anesthesia in dogs. Anesth Analg 1993; 76:1222-6.
Lam AM, Matta BF, Mayberg TS, Strebel S: Change in cerebral blood flow velocity with onset of EEG silence during inhalation anesthesia in humans: evidence of flow-metabolism coupling? J Cereb Blood Flow Metab 1995; 15:714-7.
von Knobelsdorff G, Kusagaya H, Werner C, Kochs E, Schulte am Esch J: The effects of surgical stimulation on intracranial hemodynamics. J Neurosurg Anesth 1996; 8:9-14.
Gregory GA, Eger EI II, Munson ES: The relationship between age and halothane requirement in man. Anesthesiology 1969; 30:488-91.
Stevens WC, Dolan WM, Gibbons RT, White A, Eger EI, Millar RD, de Jong RH, Elashoff RM: Minimum alveolar concentrations (MAC) of isoflurane with and without nitrous oxide in patients of various ages. Anesthesiology 1975; 42:197-200.
Katoh T, Ikeda K: The minimum alveolar concentration (MAC) of sevoflurane in humans. Anesthesiology 1987; 66:301-3.
Nagayama T, Meguro K, Inada Y: Sevoflurane MAC in the elderly. Masui 1994; 43:238-41.
Thiel A, Zickmann B, Zimmermann R, Hempelmann G: Transcranial Doppler sonography: Effects of halothane, enflurane and isoflurane on blood flow velocity in the middle cerebral artery. Br J Anaesth 1992; 68:388-93.
Bisonnette B, Leon JE: Cerebrovascular stability during isoflurane anesthesia in children. Can J Anaesth 1992; 39:128-34.
Sakabe T, Nakakimura K: Effects of anesthetic agents and other drugs on cerebral blood flow, metabolism, and intracranial pressure, Anesthesia and Neurosurgery. Edited by Cottrell JE, Smith DS. St. Louis, Mosby, 1994:149-74.
Manohar M, Parks CM: Pocrine systemic and regional organ blood flow during 1.0 and 1.5 minimum alveolar concentrations of sevoflurane anesthesia without and with 50% nitrous oxide. J Pharmacol Exp Ther 1984; 231:640-8.
Manohar M: Regional brain blood flow and cerebral cortical O sub 2 consumption during sevoflurane anesthesia in healthy isocapnic swine. J Cardiovasc Pharmacol 1986; 8:1268-75.
Scheller MS, Tateishi A, Drummond JC, Zornow MH: The effects of sevoflurane on cerebral blood flow, cerebral metabolic rate for oxygen, intracranial pressure, and the electroencephalogram are similar to those of isoflurane in the rabbit. Anesthesiology 1988; 68:548-51.
Scheller MS, Nakakimura K, Fleischer JE, Zornow MH: Cerebral effects of sevoflurane in the dog: Comparison with isoflurane and enflurane. Br J Anaesth 1990; 65:388-92.
Kitaguchi K, Ohsumi H, Kuro M, Nakajima T, Hayashi Y: Effects of sevoflurane on cerebral circulation and metabolism in patients with ischemic cerebrovascular disease. Anethesiology 1993; 79:704-9.
Cho S, Fujigaki T, Uchiyama Y, Fukusaki M, Shibata O, Sumikawa K: Effects of sevoflurane with and without nitrous oxide on human cerebral circulation. Anesthesiology 1996; 85;755-60.
Strebel S, Lam AM, Matta B, Mayberg TS, Aaslid R, Newell DW: Dynamic and static cerebral autoregulation during isoflurane, desflurane, and propofol anesthesia. Anesthesiology 1995; 83:66-76.