Direct Cerebral Vasodilatory Effects of Sevoflurane and Isoflurane 

With local ethical committee approval and written informed consent, 20 patients, aged 20–62 yr, American Society of Anesthesiologists physical status 1 or 2, requiring general anesthesia for routine spinal neurosurgical procedures, were recruited. Routine monitoring consisting of electrocardiography, noninvasive blood pressure measurement (Dinamap; Critikon, Tampa, FL), and pulse oximetry was instigated before the induction of anesthesia with propofol 2.5 mg/kg, fentanyl 2 μg/kg, and atracurium 0.5 mg/kg. After tracheal intubation, the lungs were ventilated with an air/oxygen mixture (inspired oxygen fraction 0.5) to mild hypocapnia (end-tidal carbon dioxide concentration 30–35 mmHg; measured with Capnomac Ultima; Datex, Helsinki, Finland). Esophageal temperature was measured continuously after the induction of anesthesia. Brain electrical activity was measured using a two-channel bipolar frontal montage (A-1000 EEG monitor; Aspect Medical Systems, MA), which displayed unprocessed EEG and burst suppression ratio. The burst suppression ratio is defined as the percentage of epochs in the past 63 s in which the EEG signal is considered suppressed; a burst suppression ratio of 100% implies EEG silence. The EEG electrodes were attached to the patients’ foreheads with self-adhesive, low-impedance electrodes, and the impedance was confirmed at less than 10,000 Ω before measurements were made. A 2-MHz transcranial Doppler ultrasound probe (DWL Multidop; Sipplingen, Germany) was used to measure the time-averaged mean blood flow velocity in the middle cerebral artery (V^mca). Once the best signal was obtained at a depth between 45 and 55 mm, the probe was secured in position using a specially designed frame so that the angle of insonation remained constant throughout the study period. Baseline anesthesia was maintained with a propofol infusion at a rate sufficient to achieve EEG isoelectricity (a rate of 200–300). After a 20-min period of stable propofol infusion and isoelectric EEG, V^mca, mean arterial pressure (MAP), end-tidal CO²concentration, heart rate, and esophageal temperature were measured. Patients received either sevoflurane or isoflurane in the first of two random doses adjusted for age, 12 i.e. , either 0.5 MAC sevoflurane (0.8–1.0%) or 1.5 MAC sevoflurane (2.4–3.0%), or 0.5 MAC isoflurane (0.5–0.7%) or 1.5 MAC isoflurane (1.5–2.0%). After a 15-min period of unchanged end-tidal concentration (Capnomac Ultima; Datex), the measurements were repeated. The concentration of sevoflurane or isoflurane was increased or decreased as appropriate, and after a further 15-min period of unchanged end-tidal concentration, all measurements were repeated. All measurements were performed before the start of surgery. An infusion of 0.01% phenylephrine was used as necessary to maintain MAP at baseline levels.

Results with each agent were compared using analysis of variance. The Scheffé test was used for a multiple-comparisons procedure in order to delineate where the differences were. P < 0.05 was considered to be statistically significant. All values are expressed as mean ± SD unless otherwise stated.

The main findings of the study are shown in table 1. The mean propofol infusion rate required to achieve EEG burst suppression was 258 ± 47 μg · kg⁻¹· min⁻¹. The mean weight in the sevoflurane group was greater than the isoflurane group (81 ± 14 vs.  68 ± 6 kg, respectively). There were no significant changes in end-tidal CO²concentration, heart rate, or MAP during the study period and the esophageal temperature was maintained at or above 36°C in all patients. All patients required phenylephrine (100–300 μg) to maintain MAP within 20% of baseline during 1.5 MAC anesthesia. Both sevoflurane and isoflurane significantly increased V^mcafrom baseline. This increase was significantly less during sevoflurane anesthesia (4 ± 3% at 0.5 MAC sevoflurane and 17 ± 3% at 1.5 MAC sevoflurane vs.  19 ± 3% at 0.5 MAC isoflurane and 72 ± 9% at 1.5 MAC isoflurane;P < 0.05).

This study demonstrates that, in common with other inhalational anesthetic agents, 2sevoflurane increases CBF velocity in a dose-dependent manner during propofol-induced EEG silence (maximal metabolic suppression). However, the magnitude of this increase with 0.5 and 1.5 MAC sevoflurane is significantly less than equipotent concentrations of isoflurane under similar conditions.

Transcranial Doppler (TCD) ultrasonography provides an indirect measure of CBF if the angle of insonation and vessel diameter remain constant. 13We used a specially designed frame to hold the TCD probe in position to ensure a constant angle of insonation throughout the study period. 13Available evidence suggests that the diameter of the middle cerebral artery is not significantly affected by changes in carbon dioxide tension, blood pressure, or the administration of anesthetic or vasoactive agents. 14–16Furthermore, the use of TCD ultrasonography to estimate changes in CBF if desflurane, halothane, or isoflurane are administered during propofol-induced EEG suppression has been demonstrated in a recent study. 2Therefore, it is reasonable to assume that during sevoflurane anesthesia changes in V^mcareflect changes in CBF.

Propofol produces cerebral vasoconstriction indirectly by reducing cerebral metabolism. 17In dogs, propofol has been shown to decrease EEG activity and cerebral metabolism associated with a decrease in CBF. 18Once the EEG becomes isoelectric, there is no further reduction in cerebral metabolism despite increasing doses of anesthetic. It is therefore reasonable to assume that under the baseline conditions of this study, with an isoelectric EEG, the addition of sevoflurane would cause no further reductions in cerebral metabolism. Any increases in CBF would therefore be caused by the intrinsic vasodilatory properties of sevoflurane on the cerebral vasculature.

We used the A-1000 EEG monitor to display an unprocessed EEG and to calculate the burst suppression ratio from a two-channel bipolar frontal montage. The burst suppression ratio is defined as the percentage of epochs in the past 63 s in which the EEG signal is considered suppressed; a burst suppression ratio of 100% implies EEG silence. We were able to achieve an isoelectric EEG with doses of propofol between 200 and 350 μg · kg⁻¹· min⁻¹. These doses are similar to those previously described for achieving an isoelectric EEG. 2 

We measured V^mcaat constant end-tidal carbon dioxide tensions. Changes in end-tidal carbon dioxide tension have been shown to correlate well with changes in arterial carbon dioxide tension in healthy patients. 19 

Phenylephrine was required in all patients to maintain MAP at baseline values during propofol-induced EEG silence. Phenylephrine has been shown recently not to directly affect intracranial hemodynamics in anesthetized patients. 20 

This study is very similar to the work published by Matta et al. , 2in which the direct cerebral vasodilatory effects of halothane, isoflurane, and desflurane using TCD ultrasonography in humans were demonstrated. The introduction of 0.5 MAC halothane, isoflurane, or desflurane during propofol-induced EEG silence resulted in similar increases in V^mcafor all three agents studied. However, if the concentration was increased to 1.5 MAC, the increase in V^mcafrom baseline was significantly more for isoflurane and desflurane than halothane (73 ± 4, 79 ± 7, and 48 ± 3%, respectively). This demonstrates that under conditions of maximal metabolic suppression, isoflurane and desflurane are more marked cerebral vasodilators than halothane. In this study sevoflurane increased V^mcaby 4 and 17% at 0.5 and 1.5 MAC, respectively, suggesting that the direct cerebral vasodilatory effect of sevoflurane is less than that of halothane, isoflurane, and desflurane. These findings agree with previous work published on the effect of sevoflurane on cerebral hemodynamics. Kitaguchi et al.  21investigated the cerebral vascular effects of 0.88 MAC sevoflurane in an oxygen–nitrous oxide mixture in patients with ischemic cerebral vascular disease undergoing extracranial–intracranial arterial anastomosis. CBF was determined by argon desaturation by the Kety–Schmidt method using a mass spectrometer. When the authors compared the data on CBF and cerebral metabolic requirements for oxygen obtained from this study to values in awake patients obtained in a previous study, they reported a reduction of 34% and 52% in CBF and cerebral metabolic requirements for oxygen, respectively. Similarly, previous work at our institution indicated that in the presence of a “light” propofol anesthetic, sevoflurane in humans has negligible effects on V^mcaand cerebral metabolic equivalent (CME) at 0.5 MAC and only causes small increases in V^mcaand cerebral metabolic equivalent at 1.5 MAC. 9Cerebral metabolic equivalent = CBF velocity × arteriovenous oxygen content difference and was used as a substitute for cerebral metabolic requirements for oxygen, as blood flow velocity was measured and not blood flow.

This “weak” intrinsic cerebral vasodilatory action of sevoflurane may help explain why the cerebral vasculature remains capable of responding to changes in perfusion pressure during sevoflurane anesthesia, but it is abolished during equipotent isoflurane and desflurane anesthesia because of greater cerebral vasodilatation. 10,22Furthermore, because of this weak intrinsic vasodilatory action, sevoflurane is unlikely to cause a significant increase in intracranial pressure, a feature preferred for agents used in neurosurgical anesthesia. This hypothesis is supported by the recently published data of Artru et al.  8who reported no increases in intracranial pressure during 0.5, 1.0, and 1.5 MAC sevoflurane anesthesia.

We conclude that in common with other volatile anesthetic agents, sevoflurane has an intrinsic dose-dependent cerebral vasodilatory effect. However, this effect is less than that reported for halothane, isoflurane, and desflurane at equipotent anesthetic concentrations. These findings suggest that sevoflurane has a cerebral hemodynamic profile favoring its use in neuroanesthesia.