| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
University Departments of Anesthesia and Intensive Care, Queens Medical Centre and City Hospital NHS Trust, Nottingham, United Kingdom
Address correspondence and reprint requests to Dr. Ravi P. Mahajan, University Department of Anesthesia, Queens Medical Centre, Nottingham, NG7 2UH, UK. Address e-mail to ravi.mahajan{at}nottingham.ac.uk.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Interest in the anesthetics for neuroanesthesia has increased because of the effects of these drugs on cerebral hemodynamics and ICP (7–15). In general, volatile anesthetics dilate cerebral blood vessels, increase cerebral blood volume and possibly ICP, and impair cerebral autoregulation and vascular reactivity (7–12); these effects are considered undesirable in neuroanesthesia. Conversely, studies on the effects of propofol, a commonly used IV anesthetic, have demonstrated possible avoidance of these undesirable effects with a reduction in cerebral blood volume and ICP, and preservation of cerebral autoregulation and vascular reactivity (11,13,14). However, the effects of different anesthetics on the derived values of estimated CPP and ZFP have not been described. We aimed to establish the effects of maintaining anesthesia with two different anesthetics, propofol or sevoflurane, on estimated CPP (eCPP) and ZFP as estimated using TCD.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Each patient was studied in the anesthetic room lying supine with the head resting on a pillow. Monitoring included electrocardiogram, peripheral oxygen saturation, noninvasive arterial blood pressure, and end-tidal carbon dioxide (Petco2) (Marquette Tramscope12). The left MCA was insonated via the temporal window using a 2-MHz Doppler ultrasound probe (Scimed QVL 120; Scimed, Bristol, UK). The identity of the MCA was confirmed using standard criteria, and the position of the probe was fixed with an elastic headband to maintain a constant angle of insonation (15).
Baseline measurements of heart rate, MAP, Petco2 breathing air, and MCA FV were recorded. Anesthesia was then induced with an IV target controlled infusion of propofol set at 6 µg/mL (Diprifusor; Zeneca Pharmaceuticals). The use of opiates was avoided. Rocuronium 0.6 mg/kg IV was administered and, after the onset of neuromuscular blockade, a laryngeal mask airway was inserted and the lungs were ventilated with 100% oxygen. Mechanical ventilation was adjusted to maintain Petco2 at the baseline value.
In the first 13 patients (propofol group), anesthesia was maintained by continuing target controlled infusion of propofol at 6 µg/mL. After achieving a steady-state, further recordings were made of heart rate, MAP, Petco2, and MCA FV. Mechanical ventilation was then readjusted to achieve mild hypocapnia, maintaining Petco2 at 1 kPa less than the baseline value. All of the measurements were recorded again.
In the following 10 patients (sevoflurane group), anesthesia was maintained with sevoflurane at an end-tidal concentration of 2% (approximately 1 minimum alveolar concentration). After allowing the calculated plasma concentration of propofol to decrease to <1 µg/mL, and ensuring end-tidal sevoflurane concentrations of 2% maintained for >10 min, further recordings were made of heart rate, MAP, Petco2, and MCA FV. The lungs were then hyperventilated to decrease Petco2 x 1 kPa relative to the baseline, and a third set of measurements was recorded.
All measurements were taken at steady-state which was defined as <10% variation in heart rate, MAP, MCA FV, or Petco2 over a 2-min period. It was decided to maintain MAP >75% of the baseline value with the use of IV ephedrine injection, if necessary, according to common clinical practice.
All TCD measurements during the study were taken as an average over 15 s to incorporate at least 2 full respiratory cycles. For the analysis, analog outputs of the FV maximum, determined using the outer envelope of the velocity power spectra, were used. Values of mean and diastolic flow velocities (FVmean and FVdiastolic) and simultaneously recorded MAP and diastolic blood pressure (DAP) values were taken to calculate eCPP and ZFP using the formula published by Belfort et al. (6).
|
|
From a previous study in healthy volunteers (1), we calculated that approximately 13 subjects would be required to detect a change of 20% from the baseline values of eCPP. This was at a statistical significance (
) of P < 0.05 with a power (ß) of 0.8. Thus, we recruited 13 subjects for the propofol group. The subsequent sample calculation for between-group comparison suggested that approximately 10 subjects would be required in the sevoflurane group to determine a 20% difference in eCPP or ZFP from the propofol group with a power (ß) of 0.8 and at a significance level () of P < 0.05. The data were analyzed for normality using the Anderson-Darling test. Because certain data were not distributed normally, nonparametric tests were used for within-group and between-group comparisons. The effect of anesthetic with or without hypocapnia on heart rate, MAP, MCA FV, Petco2, eCPP, and ZFP were analyzed using Friedman’s test. For within-group paired comparisons, Wilcoxon’s signed rank test was used. For between-group comparisons of the values at baseline, and at different stages of the study, the Mann-Whitney U-test was used. P < 0.05 was considered significant.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
|
|
|
The eCPP decreased significantly in patients maintained on propofol anesthesia (P = 0.002), but it did not change significantly from the baseline value in those maintained on sevoflurane anesthesia with or without hypocapnia. The eCPP values in patients from the propofol group were significantly less than the values of eCPP in patients from the sevoflurane group during anesthesia with or without hypocapnia (Fig. 1).
The ZFP increased significantly during maintenance of anesthesia in the propofol group (P = 0.018); hypocapnia caused no further increase. In contrast, ZFP decreased significantly during maintenance of anesthesia in the sevoflurane group (P < 0.05); hypocapnia caused a significant increase in ZFP from the value at anesthesia with normocapnia (P < 0.05), but even at this stage ZFP still remained significantly less than the baseline value (Fig. 2). The values of ZFP in patients from the propofol group were significantly higher than the values in patients from the sevoflurane group during anesthesia with or without hypocapnia (Fig. 2).
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The conventional understanding of CPP (calculated as MAP – ICP) assumes the existence of a Starling resistor at the level of collapsible cerebral veins. However, Weyland et al. (2) have demonstrated that vascular tone determines the effective downstream pressure in patients without intracranial hypertension. They postulated the existence of two Starling resistors in series: one at the arteriolar level and influenced by vascular tone, and the other at the level of collapsible cerebral veins and influenced by ICP. In healthy volunteers, we have shown that increases in Petco2 increased eCPP presumably by decreasing the vascular tone and thus the effective downstream pressure; decreases in Petco2 had the opposite effect (1). The findings were counterintuitive to the conventional understanding in which hypercapnia is expected to decrease CPP and hypocapnia is expected to increase it by increasing or decreasing the cerebral blood volume, and hence ICP, respectively. However, the study provided evidence that vasomotor tone has a major role in determining CPP in subjects without intracranial hypertension. In the present study, changes in eCPP and ZFP in the propofol and sevoflurane groups are similar in nature to those seen during hypocapnia and hypercapnia respectively, suggesting that the effects on vasomotor tone have a predominant role in determining effective downstream pressure.
Burton (16) described the rationale for the existence of a critical closing pressure in blood vessels. He used the Law of Laplace (Pressure = Tension/Radius) to explain that an equilibrium exists between the pressure inside a vessel, the tension in the vessel wall, and the radius of the wall. In very small vessels, premature closure can occur prior to the intraluminal pressure becoming zero, and this has been described in terms of critical closing pressure. In a vessel, at unstretched radius and at equilibrium, the active tension in the vessel wall just balances the pressure. If tone is decreased, then the pressure at which vascular closure occurs will also decrease. The concept was further supported by animal models in which techniques to alter vessel tone were used (17,18). These experimental studies on dynamic flow-pressure relationships showed that changes in vasomotor tone influenced the pressure at which the blood flow to the brain would cease (ZFP).
Using TCD, noninvasive estimation of CPP was first described by Aaslid et al. (3) who measured the mean FV (Vo) and the first harmonic of the velocity waveform (V1) and the arterial blood pressure (ABP). eCPP was then calculated as (Vo/V1) x ABP where ABP/V1 represents vascular resistance. Belfort et al. (6) modified this to (MAP – DAP)/(FVmean – FVdiastolic) to represent resistance, creating a formula for eCPP = [(MAP – DAP)/(FVmean – FVdiastolic)] x FVmean, and hence ZFP = MAP – eCPP. Other similar formulas have been described using the ratio between instantaneous MAP and FV or the change in MAP and FV to represent resistance (4,5,19). Some workers have taken a number of simultaneous points on the FV and MAP waveforms, and used regression lines between the two to estimate ZFP (2,4,20). Recently, Aaslid et al. (4) have provided data on the validation of estimation of ZFP from simultaneous recordings of MCA FV and MAP, using regression as well as intercept techniques, in a group of patients in whom ventricular fibrillation was induced. However, critical appraisal of different methods, in particular of the noninvasive methods, and the relative advantages of one method over the other in a given situation, is still needed.
In this study, as in our previous study (1), we used the method described by Belfort et al. (6) because of its simplicity and applicability to patients without invasive MAP monitoring. The baseline values of ZFP in our study showed large interindividual variability, including some negative values. Previous studies have also reported a wide range for ZFP values including some negative values (1,2,6,19,20). We believe that this is the reflection of variability in the relative pulsatilities of the MAP and FV waveforms within the study population. In theory, the estimation of ZFP from simultaneous measurements of FV and pressure during pulsatile flow in an elastic vessel, whether based on systolic blood pressure, DAP, or MAP values and FV, or on the relationship between a number of values of MAP and FV during a cardiac cycle, is likely to be confounded by factors that would affect the pulsatility of either pressure pulse wave or FV pulse wave or both. These factors are heart rate, compliance and resistance of the vascular bed distal to the point of insonation, rheological properties of the blood, and elasticity of the insonated part of the vessel (21–23). It is also worth noting that in clinical practice, for calculating ZFP, blood pressure and the FV are measured at two different sites (upper arm for blood pressure, and MCA for FV); therefore, different factors can affect their pulsatility independently. Thus, even under normal circumstances a wide range of values for ZFP is possible including the negative values. Using a bench model validation (24) of Belfort et al.’s method, we have recently shown that large variability in the values of ZFP could be produced by changing the configurations of FV and blood pressure pulsatility; however, despite the wide range of values for ZFP, its estimation remained a sensitive indicator of changes in the downstream pressure, and thus could be reliably used to assess changes in perfusion pressure and ZFP.
There are some potential limitations of the present study that merit discussion. We did not attempt to maintain MAP near baseline values in the present study. This was partly to reflect clinical practice, and partly to avoid the use of vasoactive substances that may have confounding effects on eCPP and ZFP (25). Mathematically, the influence of decreases in MAP on eCPP or ZFP is difficult to predict without an accurate prediction of the effect of hypotension on FV and its pulsatility. Physiologically, moderate hypotension would cause autoregulatory cerebral vasodilatation and any accompanying reduction in vascular tone would reduce ZFP. In the present study, the decrease in MAP was similar in both groups. Because both propofol and sevoflurane, in the doses used in the present study, preserve cerebral autoregulation (13,26), the independent effects of moderate hypotension on ZFP can be taken to be similar in both groups. Therefore, we do not believe that the presence of moderate hypotension could have influenced the main thrust of the results in this study.
Potentially, changes in ICP can affect eCPP and ZFP. This study lacks ICP measurements because these were not indicated on clinical grounds. However, in patients with no neurological disorder, propofol can be expected to decrease cerebral blood volume (and thus ICP) (14), whereas sevoflurane can be expected to increase it (12). The changes in ZFP in the present study were in the opposite direction. Thus, these changes are more likely to be caused by the effects of propofol or sevoflurane on vascular tone rather than ICP. In the present study, moderate hypocapnia resulted in an increase in ZFP in the sevoflurane group but not in the propofol group. The lack of increase in ZFP in the propofol group could have been because in these patients the vasomotor tone had already increased significantly, which precluded further increases by moderate hypocapnia. We did not randomize the sequence of hypocapnia because this would have prolonged the experimental procedure in the anesthetic room which was deemed unacceptable. However, the mean reduction in MCA FV during hypocapnia was consistent with previously reported studies in which changing Petco2 was randomized under similar experimental conditions (1). Because our study was conducted in patients with no neurological disorder, these findings cannot be taken to reflect the effects of propofol or sevoflurane in patients with neurological problems. However, they will serve as an important point of reference for further studies in such patients.
We studied the two anesthetics in tandem. Ideally, we would have randomized the patients to enter one or the other group. Because the inclusion of the sevoflurane group to the study was an afterthought, randomization could not be achieved. However, both groups were studied in similar experimental conditions, and the measurements were all objective in nature. Thus, we do not believe that lack of randomization could have had significant influence on our results.
We used 100% oxygen during anesthesia in the present study. Oxygen is a constrictor of cerebral blood vessels. It is conceivable that it may also increase cerebral vascular tone in which case its presence may tend to exaggerate the effects of propofol and attenuate the effects of sevoflurane on ZFP. However, in magnitude, the effects of 100% oxygen per se on cerebral vasculature, as assessed by changes in MCA FV and cerebral autoregulation (9), are much less marked compared with the effects of propofol or sevoflurane (13,26). Therefore, the influence of 100% oxygen in the present study could only be small quantitatively, and negligible qualitatively (as it was used in both groups). We took Petco2 during anesthesia at the same level as the baseline awake value to reflect normocapnia. The relationship between Petco2 and arterial carbon dioxide can change during anesthesia. Because both groups were treated in a similar manner, any possible confounding effects of potentially small changes in arterial carbon dioxide during anesthesia were likely to be similar in both the groups, and therefore would not be expected to have significant impact on the comparative results.
In summary, our results show that in a population without any neurological disorder, propofol and sevoflurane had opposing effects on cerebral vascular tone as estimated by eCPP and ZFP. During moderate hypotension under anesthesia, eCPP was maintained near baseline with sevoflurane whereas it decreased significantly during propofol anesthesia. The ZFP increased significantly with propofol whereas it decreases significantly with sevoflurane. A hypocapnia-induced increase in ZFP was seen during sevoflurane, but not propofol, anesthesia.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. Dinsmore Anaesthesia for elective neurosurgery Br. J. Anaesth., July 1, 2007; 99(1): 68 - 74. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Demirci, S. Unver, U. Karadeniz, Y. Cetintas, D. Kazanci, and O. Erdemli Middle Cerebral Arterial Blood Flow Velocity and Hemodynamics in Heart Surgery Asian Cardiovasc Thorac Ann, April 1, 2007; 15(2): 97 - 101. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
