JOURNAL HOME CME HOME THIS MONTH PAST ISSUES ETOC COLLECTIONS
AUTHORS REVIEWERS EDITORIAL BOARD FEEDBACK RSS HELP
A&A International Anesthesia Research Society
QUICK SEARCH:   [advanced]


     

Anesth Analg 2008; 106:899-909
© 2008 International Anesthesia Research Society
doi: 10.1213/ane.0b013e3181619685
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a colleague
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via Web of Science (2)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bonhomme, V.
Right arrow Articles by Laureys, S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bonhomme, V.
Right arrow Articles by Laureys, S.
Related Collections
Right arrow Neuroanesthesia
Right arrow Preoperative Evaluation

NEUROSURGICAL ANESTHESIOLOGY

Section Editor:
Adrian W. Gelb

The Effect of Clonidine Infusion on Distribution of Regional Cerebral Blood Flow in Volunteers

Vincent Bonhomme, MD, MSc*, Pierre Maquet, MD, PhD{dagger}{ddagger}, Christophe Phillips, PhD{dagger}, Alain Plenevaux, PhD{dagger}, Pol Hans, MD*, Andre Luxen, PhD{dagger}, Maurice Lamy, MD*, and Steven Laureys, MD, PhD{dagger}{ddagger}

From the *University Department of Anesthesia and Intensive Care Medicine, CHU de Liege and CHR de la Citadelle, and {dagger}Cyclotron Research Center, and {ddagger}Department of Neurology, University of Liege and CHU de Liege, Liege, Belgium.

Address correspondence and reprint requests to Vincent Bonhomme, MD, MSc, University Department of Anesthesia and Intensive Care Medicine, Bd du 12eme deLigne, 1, 4000 LIEGE, Belgium. Address e-mail to vincent.bonhomme{at}chu.ulg.ac.be.

Abstract

BACKGROUND: Through their action on the locus coeruleus, {alpha}2-adrenoceptor agonists induce rapidly reversible sedation while partially preserving cognitive brain functions. Our goal in this observational study was to map brain regions whose activity is modified by clonidine infusion so as to better understand its loci of action, especially in relation to sedation.

METHODS: Six ASA I–II right-handed volunteers were recruited. Electroencephalogram (EEG) was monitored continuously. After a baseline H215O activation scan, clonidine infusion was started at a rate ranging from 6 to 10 µg · kg–1 · h–1. A sequence of 11 similar scans was then performed at 8 min intervals. Plasma clonidine concentration was measured. Using statistical parametric mapping, we sought linear correlations between normalized regional cerebral blood flow (rCBF), an indicator of regional brain activity, and plasma clonidine concentration or spindle EEG activity.

RESULTS: Clonidine induced clinical sedation and EEG patterns (spindles) comparable to early stage nonrapid eye movement sleep. A significant negative linear correlation between clonidine concentration and rCBF or spindle activity was observed in the thalamus, prefrontal, orbital and parietal association cortex, posterior cingulate cortex, and precuneus.

CONCLUSIONS: The EEG patterns and decreases in rCBF of specific brain regions observed during clonidine-induced sedation are similar to those of early stage nonrapid eye movement sleep. Patterns of deactivated brain regions are also comparable to those observed during general anesthesia or vegetative state, reinforcing the hypothesis that alterations in the activity of a common network occur during these modified conscious states.

Consciousness can be operationally divided into two major components: vigilance (i.e., the level of consciousness) and awareness (i.e., the mental content: conscious perceptions, intentions, memories, emotions, inner speech and thoughts).1 Vigilance depends on the tonic excitatory influence of activating structures of the brainstem and diencephalon, which affect the activity of thalamocortical and corticocortical circuits. This persistent activation sets and maintains the necessary conditions for the content of consciousness to emerge.2

General anesthesia produces pharmacologically induced graded states of coma with loss of mental content, reduced pain, and immobility. These effects are obtained together with a decrease in the level of vigilance, which, typically, is not quickly reversible. In this respect, {alpha}2-adrenoceptor agonists such as dexmedetomidine or clonidine have been shown to induce a clinical state of sedation, which can be quickly reversed by simple physical or verbal stimulation3 without any change in the plasma concentration of the drug. The tranquility and arousability achieved with {alpha}2-adrenoceptor agonists render their use particularly convenient in clinical situations requiring sedated, comfortable, and cooperative patients.4–6 The reversibility of the {alpha}2-adrenoceptor agonist-induced sedation suggests that several basic cognitive brain functions are preserved7 and is reminiscent of the loss of consciousness consistent with normal sleep.

At sleep onset, the activating input from brainstem structures, including from the noradrenergic locus coeruleus, to the thalamus progressively decreases. Consequently, disabled thalamic neurons begin to fire in bursts within the spindle frequency range and entrain thalamocortical neurons in spindle oscillations. Spindles are a characteristic feature of early sleep state and are easily recognized on scalp electroencephalogram (EEG) recordings in humans, as waxing and waning oscillations of 12–15 Hz frequency lasting for at least a half second. As sleep deepens, slower oscillations are generated in thalamocortical circuits, which appear as slow waves on surface EEG recordings.2 Another consequence of the decrease in activating input from brainstem structures at sleep onset is a {gamma}-amino-butyric acid (GABA)-mediated inhibition of all the ascending monoaminergic, cholinergic, and orexinergic arousal nuclei that project onto the cortex, therefore decreasing cortical activity8 and contributing to the alteration of consciousness.

Functional brain imaging studies have identified the changes in regional brain activity during the above-mentioned altered conscious states, namely sleep9 and general anesthesia,10–12 as well as during vegetative state.13 Although of different origin, they share common characteristics in that respect. This leads to the hypothesis of an unconsciousness-related alteration in the activity of a common network. Alkire et al. have hypothesized that a hyperpolarization block of thalamocortical communication would, as in sleep-induced unconsciousness, be the essential common neurophysiologic mechanism underlying anesthetic-induced unconsciousness.14 This hypothesis would explain how the highly variable biochemical effects of anesthetics would ultimately converge on a final common pathway.

The clinical characteristics of the sedation induced by {alpha}2-adrenoceptor agonists are very close to those of physiological nonrapid eye movement sleep. In consideration of their ability to elicit spindling in rodents15 and humans,16 and their specific mechanism of action, namely an inhibiting effect on locus coeruleus projections,17 we designed the present study to precisely map the dose-dependent relative regional cerebral activity changes associated with clonidine infusion, while clinically and electrophysiologically characterizing the induced sedation.

METHODS

After IRB approval and informed consent, six healthy and drug-free right-handed volunteers (female/male: 5/1, age: 19–24 yr, weight: 52–70 kg, height: 168–173 cm) were recruited.

Equipment and Monitoring
On arrival in the Positron Emission Tomograph (PET)-scan ward, all volunteers were equipped with standard electrocardiogram, peripheral saturation in oxygen (Spo2), and noninvasive arterial blood pressure (NIBP) monitoring (Datex-OhmedaTM monitor, Helsinki, Finland). Electrodes were applied to allow referential recording of C3-A2 and C4-A1 EEG activity, according to the 10–20 international nomenclature. Bispectral Index® (BIS) electrodes (BIS-Sensor®, A2000 BIS® monitor, version 3.4, Aspect Medical Systems, Inc.) and A-Line® autoregressive index (AAI) electrodes (A-Line AEP electrodes, A-Line® monitor, Danmeter A/S, version 1.5) were then properly placed on the left side of the forehead, and earphones securely fixed on the subject's ears using plastic tape (OpSiteTM FlexifixTM, Smith & Nephew, Brussels, Belgium). The AAI is an index ranging between 0 and 100. It is derived from middle latency auditory-evoked potentials and has been validated as a measure of the level of sedation.18 Volunteers received the auditory stimulation of the A-Line® device (bilateral click stimulus of 70-dB intensity and 2 ms duration) continuously from that moment to the end of the study. Two 18-gauge IV catheters were inserted into a vein of the left and right elbow. The left catheter received continuous saline infusion, labeled water (H215O), and clonidine infusion. The right one was for blood sampling. Each volunteer was comfortably placed on the PET-scan tray and the head was stabilized by a thermoplastic facemask secured to the head holder (Truscan Imaging, Annapolis, MA). Subjects were breathing through a plastic facemask delivering oxygen at a rate of 10 L · /min–1.

Data Acquisition
The sequence of data acquisition was the same for each volunteer. Throughout the procedure, BIS, AAI, NIBP, heart rate (HR), and Spo2 were continuously recorded using a laptop computer connected to the concerned monitors and using the RugloopII © monitor-only software (Demed, Temse, Belgium). Sampling rate of RugloopII © software was 1 Hz for AAI and BIS, and 0.2 Hz for HR, NIBP, and Spo2. EEG was also continuously recorded on a separate computer, at a sampling rate of 500 Hz, using the NeuroscanTM software (Version 3.0, Compumedics, Hamburg, Germany). Direct visual observation was maintained at all times. All subsequent events took place in a quiet, low-light environment, and subjects were asked to keep their eyes closed at all times.

A transmission scan was first performed to measure attenuation and to allow subsequent corrections. After the end of the transmission scan, a sequence of 12 emission scans (activation scans) was acquired, at 8-min intervals each, in 3-dimensional mode and using a CTI 951 16/32 scanner (Siemens, Erlangen, Germany). Each scan consisted of two frames: a 30-s background frame and a 90-s frame. A slow IV H215O infusion was started immediately before the second frame to observe the head curve rising within the first 10 s of this frame. Six to eight millicuries (222–296 MBq) were injected for each scan, in 10 mL saline, over a period of 20 s. The infusion was totally automated to alleviate disturbance of the volunteer during the scanning period. Data were reconstructed using a Hanning filter (cutoff frequency: 0.5 cycle per pixel) and corrected for attenuation and background activity. The images were normalized for differences in global cerebral blood flow (CBF) by means of ratio normalization: i.e., the count at each voxel was divided by the mean counts calculated across all brain voxels. A high resolution (voxel size: 0.96 x 0.96 x 1.35 mm) T1-weighted structural magnetic resonance imaging scan was obtained for each subject on a 3 T imager (Allegra, Siemens) a few days after the PET session.

The first activation scan was performed in the absence of clonidine. The infusion of clonidine was started immediately after the first activation scan in each subject at a constant rate varying between 6 and 10 µg · kg–1 · h–1. The rate was chosen prior to the session according to a randomization list. The infusion of clonidine was then stopped randomly during the remaining 11 scan sequence, again according to the randomization list. These randomizations were instituted to reduce the risk of an order effect. The infusion regimen for each volunteer is detailed in Table 1. Response to verbal command was evaluated 1 min before each scan: the subjects were asked to squeeze the hand of the observer. The awareness testing was short: subjects were asked to squeeze the hand only once, using a loud clear voice. If no response was obtained, no supplementary test of awareness was performed until the following scan. A 10-mL blood sample was drawn immediately after each scan, centrifuged, and frozen at –20°C to allow post hoc plasma clonidine concentration measurements.


View this table:
[in this window]
[in a new window]

  		Table 1. Characteristics of Clonidine Infusion for Each Volunteer

 

Clonidine Plasma Concentration Measurement
Plasma clonidine concentrations were measured using the high performance liquid chromatography (HPCL) liquid chromatography (LC) tandem mass spectrometry (MS/MS) technique described by Pelzer et al.19 Clonidine hydrochloride and human blank serum were from Sigma-Aldrich, clonidine-d4 hydrochloride was obtained from Euriso-Top (France), and all other chemicals were from Merck. The approach consisted of clonidine and internal standard clonidine-d4 extractions from alkalized serum with ethyl ether followed by evaporation–reconstitution process. Clonidine was quantified with high performance liquid chromatography using an MS/MS detector through monitoring of m/z 230 {Pi} 213 transitions for clonidine and of m/z 236 {Pi} 219 transitions for internal standard clonidine-d4. A detailed description of this method is provided in the Appendix.

Data Analysis and Statistics
For all statistical analyses, a two-tailed P ≤ 0.05 was considered significant, unless otherwise indicated. Normality of distribution was checked when required.

BIS, AAI, and EEG Analysis
BIS and AAI recorded values were averaged over the 2-min scanning period, leading to 12 mean values for each subject. A least square linear regression was calculated between plasma clonidine concentration and BIS or AAI values.

Scalp EEG recordings were first analyzed visually by a certified electrophysiologist to detect specific sleep patterns. In addition, data were submitted to a time–frequency analysis using a wavelet transformation of the signal based on the method described by Tallon-Baudry et al.20 This method provides a better compromise between time and frequency resolutions than conventional Fourier transforms and is suited to detect the occurrence of spindles (waxing and waning oscillations of 12–15 Hz frequency and at least half a second duration). The signal recorded in C3-A2 was convoluted with complex Morlet's wavelets.21 A wavelet family was defined for frequencies ranging from 7 to 15 Hz in 0.5 Hz steps, based on a fixed number of oscillations (Noscillations = 32). At 7 Hz, this lead to a wavelet duration of 2286 ms and at 15 Hz, to a duration of 1067 ms. Therefore, the time resolution increased with frequency. The time-varying energy of the signal in a given frequency band was computed as the norm of the wavelet coefficients. For each frequency band of 0.5 Hz, the energy at any time point was divided by the variance of the energy computed before clonidine infusion, thereby providing a Z score for each time–frequency point. A time–frequency map was then constructed (Fig. 3), showing the evolution of the complex and fluctuating EEG activity over time. This map displays the Z scores of each frequency band as a function of time in the form of a color scale. The time–frequency map was thresholded to 3.09, corresponding to a significant change in energy, when compared with baseline levels, at P < 0.001.


Figure 333
View larger version (75K):
[in this window]
[in a new window]

  		Figure 3. Time–frequency analysis of the electroencephalogram (EEG) during the study period. A time–frequency map is presented for each volunteer. Those maps are thresholded to 3.09, corresponding to a significant change in energy, when compared with baseline levels, at P < 0.001. Red arrows correspond to the start of clonidine infusion and the blue arrows correspond to the time at which clonidine infusion was stopped. Blue lines indicate the scanning periods. Note the significant increase of energy in lower frequency bands (7–8 Hz) as well as in the spindle frequency range (12–15 Hz) after the infusion of clonidine was started. The increase in spindle activity can be observed in all subjects, although only moderately in subjects 2 and 3. Yellow rectangles are superimposed to highlight the frequency maps obtained when subjects did not respond to verbal command. Note that these periods are characterized by an intense activity in lower and/or higher frequency bands compared with the remaining of the recording.

 

PET Data Analysis
The analysis of PET data was performed using the Statistical Parametric Mapping software (Version 2, SPM2; Wellcome Department of Cognitive Neurology, Institute of Neurology, London, United Kingdom),22 implemented in Matlab © software (version 7.0.1., Mathworks Inc., Natick, USA). A simplified description of the method can be found in the review paper by Maquet.9 Data from each subject were first realigned using a least-squares approach and the first scan as a reference to correct for head movements of the subject during the experimental session. PET data were then coregistered to individual T1-weighted magnetic resonance imaging scans. After realignment, all images were transformed into a standard space (according to the atlas of Talairach and Tournoux23), to correct for intersubject differences in size and shape of the brain. All these spatial transformations allowed fitting all images into the same standardized space and were necessary for the voxel-based statistical analysis and the precise anatomical identification of the concerned brain regions (a particular voxel represents the same brain area in all scans of the same subject and in the different experimental subjects). Finally, the images were smoothed using a 16-mm full width at half-maximum isotropic Gaussian kernel to increase the signal-to-noise ratio.

Statistical analysis of regional CBF (rCBF) was performed using normalized rCBF of the six subjects, scanned 12 times each. Assessing the significance of the relationship between the measured plasma concentration of clonidine and the normalized rCBF (i.e., their linear regression) was performed by means of an analysis of covariance. The main effect was the subject's effect, and plasma concentration of clonidine was the covariate. The data set consisted of 72 rCBF volumes (12 volumes x 6 subjects). The parameter of interest was the slope of the relationship between the plasma concentration of clonidine and normalized rCBF. The subject's effect was first removed and a regression t-statistic map was calculated thereafter. An estimate of the slope and its standard deviation (sd) were obtained by least-squares fitting of the model (analysis of covariance) at each voxel. Seventy-two values of covariate were used, corresponding to the 72 volumes in the dataset. The degrees of freedom of the sd were increased from 59 (72-12-1) by pooling the sd across all voxels, so that the distribution of the t-statistic map was normal. The resulting t-statistic map tested whether, at a given voxel, the slope of the regression was significantly different from zero. Two types of map were therefore constructed: the first one indicating brain regions where there was a negative correlation between normalized rCBF and clonidine concentration, and the other indicating brain regions with a positive correlation. A significant negative correlation indicates a decrease in relative regional brain activity as a function of clonidine concentration, and a positive one, an increase. The regression t-maps were transformed to the unit normal distribution (Z distribution). The presence of a significant peak was tested by a method based on the three-dimensional Gaussian random-field theory, which corrects for the multiple comparisons involved in searching across a volume. Results were thresholded for significance at false-discovery rate corrected P < 0.01.24

In a subsequent independent analysis, a similar method was applied to identify negative and positive linear correlations with spindle band power (12–15 Hz) and {delta} power (1–4 Hz) calculated for each PET scan. Here, results were interpreted at uncorrected P < 0.05 based on the previously identified brain regions that showed significant linear correlations with clonidine concentration.

RESULTS

Plasma Concentration of Clonidine
The measured plasma concentrations of clonidine are presented in Figure 1. For technical reasons, one measure was not available in three volunteers, leading to three missing values. Those values were replaced by the mean between the preceding and following measure in each individual subject and served for subsequent data analysis. Measured concentrations ranged between 0 and 2.64 ng · mL–1.


Figure 133
View larger version (15K):
[in this window]
[in a new window]

  		Figure 1. Measured plasma concentrations of clonidine in nanogram per milliliter at each scan. Lower part provides individual values for each subject. The dashed line is interrupted when the value is a missing value replaced by the mean between preceding and following measured concentration. Clonidine infusion always started between scan 1 and scan 2. Infusion rate is provided for each subject (µg · kg–1 · h–1). Moments in the scan sequence at which the clonidine infusion was interrupted are indicated as large open squares. The time points at which subjects did not respond to verbal command are indicated as large open circles. Those time points corresponded to the end of clonidine infusion at scan 10 for subject 3 and at scan 8 for subject 5. At those two points, large open circles are superimposed on the large open squares. Upper part of the figure provides the overall mean ± sd for all subjects as a function of the scan sequence.

 

Vital Signs
Hemodynamic variables as a function of plasma clonidine concentrations are presented in Figure 2. Mean arterial blood pressure (MAP) remained remarkably stable in each volunteer and we were not able to show any dose–response relationship between clonidine concentration and MAP. HR significantly and linearly decreased as clonidine concentration increased (slope = –4.70, F(1, 70) = 44.06, P < 0.001; constant = 54.53; r2 = 0.39, t(70) = 6.64, P < 0.001). The lowest MAP and HR values observed were 65 and 40, respectively. Spo2 remained perfectly stable during the whole study at a value very close to 99%. No ventilation disturbances, such as airway obstruction or decrease in respiratory rate, were observed at any time.


Figure 233
View larger version (9K):
[in this window]
[in a new window]

  		Figure 2. Mean arterial blood pressure (MBP) (mm Hg; closed squares) and heart rate (HR) (bpm; open circles) as a function of plasma clonidine concentration (ng · mL–1). Results of least square linear regression are also provided.

 

Response to Verbal Command, EEG, BIS, and AAI
The total time spent by our volunteers in the PET-scanner from the transmission scan to the end of the 12th activation scan was [mean (sd)] 147 min.9 Throughout the study, volunteers remained easily arousable, as all of them responded adequately to the command of the observer (squeeze the hand) with few exceptions: Subject 2 did not respond to the command at scan 10, subject 5 at scan 6, and subject 6 at scan 5 and 8. Those subjects did respond adequately at all other scans. Corresponding plasma concentrations of clonidine at the time of no response were 1.91, 1.02, 2.01, and 1.36 ng/mL, respectively. None of the volunteers reported any dreams or hallucinations once having completed the study.

Visual analysis of the EEG clearly identified recurrent spindles in 5 of the 6 volunteers during clonidine infusion. Time–frequency domain analysis typically revealed that the energy in the lower frequency range (7–8 Hz) as well as in the spindle frequency range (12–15 Hz) increased after the infusion of clonidine started. The increase in spindle activity was observed in all subjects, although only moderately in two subjects (subject 2 and 3, Fig. 3).

The relationship between plasma concentration of clonidine and BIS or AAI is shown in Figure 4. There was a significant negative linear correlation between clonidine concentration and BIS value (slope = –8.45, F(1, 70) = 25.06, P < 0.001, constant = 86.30, r2 = 0.26, t(70) = 5.01, P < 0.001). Because of high impedance values and poor signal quality, AAI recordings were available only in four volunteers. There was no evident dose–response relationship between clonidine concentration and AAI value.


Figure 433
View larger version (9K):
[in this window]
[in a new window]

  		Figure 4. Bispectral index (BIS) (closed circles) and autoregressive index (AAI) value as a function of plasma clonidine concentration (ng · mL–1). Results of least square linear regression are also provided.

 

Relationship Between rCBF and Plasma Clonidine Concentration
There was a significant negative linear correlation between plasma clonidine concentration and rCBF in several brain regions including prefrontal, orbital and parietal association cortex, posterior cingulate, precuneus and thalamus/brain stem, as displayed in Figure 5A. A significant positive linear correlation was found in the upper part of the temporal lobe and the fusiform gyrus, bilaterally (Fig. 5C). A detailed description of peaks can be found in Table 2.


Figure 533
View larger version (63K):
[in this window]
[in a new window]

  		Figure 5. Brain areas showing a negative (A; shown in blue) and positive (C; shown in orange) linear correlation between plasma clonidine concentration and regional cerebral blood flow (rCBF) projected on a spatially normalized magnetic resonance imaging (MRI) and thresholded at false discovery rate corrected P < 0.01.23,24 Section B shows negative correlations with spindle frequency band (12–15 Hz) power (illustrated at uncorrected P < 0.05). Note that increased spindle activity correlates with decreases in rCBF in similar brain areas where clonidine induced rCBF reduction.

 

View this table:
[in this window]
[in a new window]

  		Table 2. Stereotaxic Localization of Peak Voxels Identified by the Regression Analysis

 

Relationship Between rCBF and EEG Power
A negative linear correlation was observed between the EEG power spectrum in the spindle frequency band and rCBF in all of the identified areas showing negative correlations with plasma clonidine concentration (i.e., prefrontal, orbital and parietal association cortex, posterior cingulate, precuneus and thalamus/brain stem, as displayed in Fig. 5B). No significant positive linear correlations with EEG spindle power were identified. No significant negative or positive correlations were identified with slow EEG changes (i.e., {delta} power).

DISCUSSION

In the present study, we demonstrate that the decrease in mental content and level of vigilance induced by clonidine, as well as the clonidine-induced occurrence of spindles in the EEG, are associated with a dose-dependent decrease in the relative activity, as indicated by rCBF, of specific brain regions. Clinical and electrophysiological characteristics of the induced sedation state, as well as the distribution of rCBF changes, are very similar to those of light non-REM sleep. Finally, a highly significant dose-dependent increase in rCBF was noted in temporal regions, which could account for some clinical effects of clonidine.

Characteristics of Clonidine-Induced Sedation
Plasma concentrations of clonidine in the range of 2 ng/mL induced tranquility and arousability in our volunteers, who remained almost continuously responsive to verbal command. This is in accordance with the findings of Hall et al.,7 who described the same moderate but significant sedation at concentrations in the same range which were associated with memory and motor skill impairments. Their volunteers always remained easily arousable and able to perform cognitive testing. We did not test memory impairment, motor skill, or cognition in an attempt to be able to detect changes in cerebral activity from a state as close as possible to a baseline resting state. However, once having completed the study, all volunteers reported that they lost their ability to perceive elapsed time during the study, and that time spent in the scanner had seemed short. Considering the results of Hall et al. and our data, we can reasonably assume that we reached our objective: mild but significant and clinically relevant sedation, arousability, and poor cognitive impairment. It is also noteworthy that the clinical state attained by our volunteers corresponded, with few exceptions, to the definition of a sleep state or, in other words, a lying and resting state, with eyes closed and reduced responses to external stimuli. Using higher doses of clonidine would probably have impaired the level of consciousness to a greater extent and rendered volunteers unresponsive to verbal stimulation more often than we observed.

The clinically relevant level of sedation is further confirmed by the dose-dependent decrease in BIS. The lowest observed values were in the range of 50, which is again in accordance with the results of Hall et al. One could argue that the significant linear correlation between clonidine concentration and BIS was calculated on measurements that were not independent, and that there could be a time series effect. However, a Durbin-Watson statistic performed on those data was 1.53, indicating a minor positive serial correlation, and therefore a minor effect of time series on that correlation. On the contrary, there was no evident relationship between AAI value and clonidine concentrations. Although this may have been due to low power in detecting the effect of clonidine on this auditory evoked potential-derived index, an explanation could be that {alpha}2-adrenoceptor agonists poorly affect the Pa-Nb components of auditory evoked responses,25,26 which are essential elements of AAI calculation.27 In this respect, BIS seems to be more sensitive to {alpha}2-adrenoceptor agonist-induced sedation than AAI.

Importantly, clonidine infusion also affected spontaneous EEG activity, as spindles, the hallmark of sleep onset, were observed in most of our volunteers. This is in accordance with the results of Bischoff et al.,16 and could be used to support the hypothesis of Nelson et al.17 that {alpha}2-adrenoceptor agonist-induced sedation occurs through the promotion of a sleep-mode pattern in non-REM sleep functional assemblies. However, other drugs, such as propofol28 or midazolam,29 can induce EEG spindling, although their mechanism of action is substantially different than that of clonidine.

Clonidine-Related Decreases in Relative rCBF
Many studies of different drugs and mechanisms of unconsciousness, namely sleep,9 general anesthesia,10,12,30 and the vegetative state,31,32 seem to show suppression of a common network. This network is composed of corticothalamocortical neural loops involving the ascending reticulothalamic activating system, frontal and parietal association cortex, the posterior cingulate cortex and the precuneus. We further tested whether this finding will generalize by observing changes of rCBF, an indicator of regional activity, caused by a drug that should have a known mechanism of action and one that should not be like most of the other anesthetics that affect GABA.8 The present study demonstrates that this network is also involved in the alteration of consciousness induced by {alpha}2-adrenoceptor agonists. It thus supports the hypothesis of a switch role for the thalamus during anesthetic-induced unconsciousness, a hypothesis proposed by Alkire et al.14 In addition, regions that show decreased activity during the above-mentioned altered states of consciousness are very similar to those involved in the default mode network of the wandering mind hypothesized by Mason et al.33 Progressively shutting down the thalamic switch could progressively shut down this default mode network before total loss of consciousness. Enhanced GABAergic inhibition would be the final common pathway of reduced cortical activity.8 This hypothesis remains compatible with that of Nelson et al.,17 which postulates that {alpha}2-adrenoceptor agonist sedation occurs through the promotion of a functional sleep pattern in endogenous neural networks. These networks would inhibit locus coeruleus neurons, which, in turn, would disinhibit other GABA-releasing subcortical nuclei. Through these effects, {alpha}2-adrenoceptor agonists would act on specific neural pathways to induce a state that is electrophysiologically and functionally similar to stage I and II non-REM sleep.

Differences in up or down regulation of the activity of thalamic and basal forebrain nuclei could be responsible for differential modulation of cortical activity, explaining the observed differences in the qualitative aspects of the alteration of consciousness during sleep, general anesthesia, vegetative state, and {alpha}2-adrenoceptor agonist-induced sedation. Each situation would be induced by a cascade of events, with different effects on various circuits but leading to a final common pattern of deactivations. This hypothesis is in accordance with the results of Coull et al., who demonstrated that the alteration in performance of a target detection task during dexmedetomidine administration can be counteracted by presentation of a loud white noise, as opposed to midazolam-induced sedation, and that this effect is associated with an increase in the activity of the left pulvinar nucleus of the thalamus.34

Beside the study of Coull et al.,34 other studies have investigated the effect of {alpha}2-agonists on global and/or rCBF. Prielipp et al. have demonstrated that an infusion of dexmedetomidine reduces the absolute CBF globally and in selected cortical regions of interest.35 However, these authors did not perform any statistical parametric mapping of the dose–response relationship between the amount of dexmedetomidine and relative rCBF, rendering comparisons with our study difficult. Fu et al. have compared the distribution of rCBF between depressed patients and nondepressed controls when administering IV clonidine.36 In terms of rCBF decreases, they found thalamic, frontal cortex and angular gyrus effects. However, the infused dose of clonidine was much lower than ours and their analysis did not evaluate concentration–effect correlations.

Clonidine-Related Increases in Relative rCBF
In addition to rCBF decreases in specific brain regions, clonidine infusion induced relative increases in the rCBF elsewhere: the upper part of the temporal cortex and the fusiform gyrus, bilaterally. The relative increase in bilateral superior temporal cortex rCBF, also observed by Fu et al.,36 was strong and topographically large. Interpretation of this increase must consider that all volunteers were continuously receiving auditory stimulation to allow measuring of AAI. However, the concerned region was much larger than the one thought to be involved in the processing of continuous neutral noise.37 During propofol sedation, when compared with baseline, a larger region of statistical significance in the temporal lobe in response to auditory stimulation has already been observed by other authors.38,39 This is possibly related to a more uniform response to stimulation, or to a change in background CBF. Therefore, the activation in the temporal lobe may simply be an artifact of the auditory stimulation applied to our volunteers.

It must also be mentioned that clonidine induces hallucinations,40–42 and that temporal cortex is involved in inner speech generation.43 Temporal activation by clonidine could therefore be related to the ability of this medication to favor auditory hallucinations, although none of the recruited volunteers reported having experienced such a phenomenon. Fusiform gyrus is involved in several cognitive functions: face recognition,44,45 phonological decoding and processing of words,46,47 spelling and rhyming judgments on visually presented words,48 and semantic processing.49,50 The increase in rCBF in that region during clonidine infusion and continuous neutral sound is not easy to interpret but could be related to facilitation by clonidine of the activity of high-order sensory processing brain regions, similarly to facilitation of upper temporal lobe activity.

Potential Limitations of the Results of the Study
The potential limitations of the present study include possible deregulation of cerebral hemodynamics by clonidine, verbal stimulation immediately before each scan, occurrence of natural sleep, and/or an order effect.

We did not measure absolute CBF during the study. This would have required the insertion of an arterial line in our volunteers. Clonidine had few systemic hemodynamic effects, mostly a moderate dose-dependent decrease in HR, and probably did not considerably influence cardiac output. It decreased global absolute CBF, as already demonstrated for {alpha}2-adrenoceptor agonists, through a direct vasoconstrictive effect on cerebral vasculature, a decrease in global cerebral metabolism, or both.35 Clonidine also mildly attenuates cerebral CO2 reactivity, but does not alter CBF autoregulation.51 We did not measure arterial CO2 partial pressure, which might have increased as a result of sedation. Therefore, we cannot exclude that the effects of clonidine on rCBF are the result of cerebral hemodynamic changes unrelated to regional synaptic activity. In our opinion, this is unlikely. Indeed, the cerebral hemodynamic changes we just mentioned were described on a global scale. To our knowledge, there are no data on regional differences in CO2 reactivity, distribution of {alpha}-adrenergic receptors on brain vessels, or CBF regulation in response to a change in MAP.

We tested the arousability of our volunteers immediately before each scan. This might have influenced cerebral activity in a way that might have masked some effects of clonidine. However, the stimulation was very short, and volunteers recovered their preceding level of sedation rapidly, as attested by the negative correlation between plasma concentration of clonidine and BIS values measured during the scans.

We cannot exclude that our volunteers experienced naturally occurring sleep during the study, independently from clonidine infusion. However, the clinical state was clearly different before when compared with during the infusion of clonidine, although volunteers experienced prolonged quiet periods during the initial part of the study (i.e., during the transmission scan in the absence of clonidine). Furthermore, the correlations we describe here are those observed between rCBF and plasma clonidine concentration. Should natural sleep have occurred, it would have been very difficult to distinguish between one phenomenon and the other, either clinically or electrophysiologically.

Finally, randomization was constrained by the fact that clonidine has a very long elimination half-life, approaching 8 h.52 We were therefore unable to alternate the presence of clonidine with clonidine-free conditions, due to unaffordable study length for the volunteers. Consequently, we cannot exclude the occurrence of an order effect. We have partially circumvented this problem by randomizing clonidine infusion duration and rate between subjects.

CONCLUSION

We conclude that the clonidine-induced altered state of vigilance with partially preserved cognitive functions is associated with specific changes in regional cerebral activity. The pattern of regional deactivation and the clinical and electrophysiological characteristics of the induced sedation are very close to those of physiological early stage non-REM sleep. Additional similarities with other altered states of vigilance and mental content, such as general anesthesia and vegetative state, reinforce the hypothesis of an alteration in the activity of a network common to all these states, and support the hypothesis of a switch role for the thalamus.

APPENDIX: CLONIDINE PLASMA CONCENTRATION MEASUREMENT TECHNIQUE

Surveyor autosampler and Surveyor MS pump (ThermoElectron Corp.) were used with a Betasil silica 5 µm column (50 x 3.0 mm, ThermoElectron Corp.). The mobile phase was acetonitrile-water-formic acid (80:20:1, v/v/v) at a flow rate of 400 µL · /min–1. Injection volume was 40 µL for a run time of 3 min. Finnigan TSQ7000 mass spectrometer (ThermoElectron Corp.) was operated in MRM (multiple reaction mode) with ESI+ source. The following settings were used: ionspray needle 4.5 KV, heated capillary 350°C, sheat gas 80, auxiliary gas 40, collision cell offset –30 V, collision cell pressure 4 mT, gain 1500 V. The transitions (precursor to product) monitored were m/z 230->213 and m/z 236 -> 219 with a window of 1.4 UM.

Calibration standards were prepared from stock-spiking solutions of clonidine in methanol at concentrations of 10, 40, 170, 630, 840, and 2000 pg/mL through the following procedure. To a glass tube (16 x 125 mm) were added blank serum (1.0 mL), the desired clonidine amount in methanol (total volume of methanol was 100 µL for each concentration), internal standard clonidine-d4 (50 µL of a 10 ng/mL solution in methanol), and ammonium hydroxide (50 µL, 1 N). The tube was vortex-mixed for 30 s. Ethyl ether (4 mL) was then added and the mixture was vortex-mixed for 3 min. The two layers were separated with centrifugation (2500g, 4°C, 5 min) and the aqueous (bottom) layer was rapidly frozen in a dry ice/methanol bath. The organic layer was transferred into a clean glass tube and evaporated to dryness with a stream of nitrogen. The samples were reconstituted with 0.1% formic acid in acetonitrile (200 µL) and vortex-mixing for 2 min.

Serum (1.0 mL), methanol (100 µL), internal standard clonidine-d4 (50 µL of a 10 ng/mL solution in methanol), and ammonium hydroxide (50 µL, 1 N) (unknowns) were treated as described for standards.

XCalibur 1.1 was used for data acquisition and processing and Microsoft Excel was used for data analysis. A weighted "square root" linear regression was used to generate calibration curves from standards and processing of unknowns.

Footnotes

Accepted for publication November 8, 2007.

REFERENCES

  1. Niedermeyer E. Consciousness: function and definition. Clin Electroencephalogr 1994;25:86–93[Web of Science][Medline]
  2. Steriade M. Sleep, epilepsy and thalamic reticular inhibitory neurons. Trends Neurosci 2005;28:317–24[Web of Science][Medline]
  3. Mantz J. Dexmedetomidine. Drugs Today (Barc) 1999;35:151–7[Medline]
  4. Mack PF, Perrine K, Kobylarz E, Schwartz TH, Lien CA. Dexmedetomidine and neurocognitive testing in awake craniotomy. J Neurosurg Anesthesiol 2004;16:20–5[Web of Science][Medline]
  5. Arain SR, Ebert TJ. The efficacy, side effects, and recovery characteristics of dexmedetomidine versus propofol when used for intraoperative sedation. Anesth Analg 2002;95:461–6[Abstract/Free Full Text]
  6. Hall JE, Uhrich TD, Barney JA, Arain SR, Ebert TJ. Sedative, amnestic, and analgesic properties of small-dose dexmedetomidine infusions. Anesth Analg 2000;90:699–705[Abstract/Free Full Text]
  7. Hall JE, Uhrich TD, Ebert TJ. Sedative, analgesic and cognitive effects of clonidine infusions in humans. Br J Anaesth 2001;86:5–11[Abstract/Free Full Text]
  8. Nelson LE, Guo TZ, Lu J, Saper CB, Franks NP, Maze M. The sedative component of anesthesia is mediated by GABA(A) receptors in an endogenous sleep pathway. Nat Neurosci 2002;5:979–84[Web of Science][Medline]
  9. Maquet P. Functional neuroimaging of normal human sleep by positron emission tomography. J Sleep Res 2000;9:207–31[Web of Science][Medline]
  10. Fiset P, Paus T, Daloze T, Plourde G, Meuret P, Bonhomme V, Hajj-Ali N, Backman SB, Evans AC. Brain mechanisms of propofol-induced loss of consciousness in humans: a positron emission tomographic study. J Neurosci 1999;19:5506–13[Abstract/Free Full Text]
  11. Alkire MT, Pomfrett CJ, Haier RJ, Gianzero MV, Chan CM, Jacobsen BP, Fallon JH. Functional brain imaging during anesthesia in humans: effects of halothane on global and regional cerebral glucose metabolism. Anesthesiology 1999;90:701–9[Web of Science][Medline]
  12. Kaisti KK, Langsjo JW, Aalto S, Oikonen V, Sipila H, Teras M, Hinkka S, Metsahonkala L, Scheinin H. Effects of sevoflurane, propofol, and adjunct nitrous oxide on regional cerebral blood flow, oxygen consumption, and blood volume in humans. Anesthesiology 2003;99:603–13[Web of Science][Medline]
  13. Laureys S. Functional neuroimaging in the vegetative state. Neuro Rehabilitation 2004;19:335–41[Medline]
  14. Alkire MT, Haier RJ, Fallon JH. Toward a unified theory of narcosis: brain imaging evidence for a thalamocortical switch as the neurophysiologic basis of anesthetic-induced unconsciousness. Conscious Cogn 2000;9:370–86[Web of Science][Medline]
  15. Buzsaki G. The thalamic clock: emergent network properties. Neuroscience 1991;41:351–64[Web of Science][Medline]
  16. Bischoff P, Scharein E, Schmidt GN, von Knobelsdorff G, Bromm B, Esch JS. Topography of clonidine-induced electroencephalographic changes evaluated by principal component analysis. Anesthesiology 2000;92:1545–52[Web of Science][Medline]
  17. Nelson LE, Lu J, Guo T, Saper CB, Franks NP, Maze M. The alpha2-adrenoceptor agonist dexmedetomidine converges on an endogenous sleep-promoting pathway to exert its sedative effects. Anesthesiology 2003;98:428–36[Web of Science][Medline]
  18. Struys MM, Vereecke H, Moerman A, Jensen EW, Verhaeghen D, De Neve N, Dumortier FJ, Mortier EP. Ability of the bispectral index, autoregressive modelling with exogenous input-derived auditory evoked potentials, and predicted propofol concentrations to measure patient responsiveness during anesthesia with propofol and remifentanil. Anesthesiology 2003;99:802–12[Web of Science][Medline]
  19. Pelzer M, Addison T, Li W, Jiang X, Weng N. Development and validation of a liquid chromatography-tandem mass spectrometry method, using silica column and aqueous-organic mobile phase, for the analysis of clonidine as low as 10 pg/mL in human serum. J Liq Chrom Rel Technol 2002;25:1019–32
  20. Tallon-Baudry C, Bertrand O, Delpuech C, Pernier J. Stimulus specificity of phase-locked and non-phase-locked 40 Hz visual responses in human. J Neurosci 1996;16:4240–9[Abstract/Free Full Text]
  21. Kronland-Martinet R, Morlet J, and Grossmann A. Analysis of sound patterns through wavelet transforms. Int J Pattern Recognit Artif Intelligence 1987;1:273–302
  22. Frackowiak RS, Friston KJ, Frith C, Doaln R, Price CJ, Zeki S, Ashburner J, Penny WD. Human Brain Function. 2 ed. San Diego: Academic Press, 2003
  23. Talairach J, Tournoux P. Co-Planar Stereotaxic Atlas of the Human Brain. 3-Dimensional Proportional System: An Approach to Cerebral Imaging. New York: Thieme Medical Publishers, 1988
  24. Genovese CR, Lazar NA, Nichols T. Thresholding of statistical maps in functional neuroimaging using the false discovery rate. Neuroimage 2002;15:870–8[Web of Science][Medline]
  25. Thornton C, Lucas MA, Newton DE, Dore CJ, Jones RM. Effects of dexmedetomidine on isoflurane requirements in healthy volunteers. 2: Auditory and somatosensory evoked responses. Br J Anaesth 1999;83:381–6[Abstract/Free Full Text]
  26. Haenggi M, Ypparila H, Hauser K, Caviezel C, Korhonen I, Takala J, Jakob SM. The effects of dexmedetomidine/remifentanil and midazolam/remifentanil on auditory-evoked potentials and electroencephalogram at light-to-moderate sedation levels in healthy subjects. Anesth Analg 2006;103:1163–9[Abstract/Free Full Text]
  27. Jensen EW, Lindholm P, Henneberg SW. Autoregressive modeling with exogenous input of middle-latency auditory-evoked potentials to measure rapid changes in depth of anesthesia. Methods Inf Med 1996;35:256–60[Web of Science][Medline]
  28. Wolter S, Friedel C, Bohler K, Hartmann U, Kox WJ, Hensel M. Presence of 14Hz spindle oscillations in the human EEG during deep anesthesia. Clin Neurophysiol 2006;117:157–68[Web of Science][Medline]
  29. Veselis RA, Reinsel RA, Beattie BJ, Mawlawi OR, Feshchenko VA, DiResta GR, Larson SM, Blasberg RG. Midazolam changes cerebral blood flow in discrete brain regions: an H2(15)O positron emission tomography study. Anesthesiology 1997;87:1106–17[Web of Science][Medline]
  30. White NS, Alkire MT. Impaired thalamocortical connectivity in humans during general-anesthetic-induced unconsciousness. Neuroimage 2003;19:402–11[Web of Science][Medline]
  31. Laureys S, Lemaire C, Maquet P, Phillips C, Franck G. Cerebral metabolism during vegetative state and after recovery to consciousness. J Neurol Neurosurg Psychiatry 1999;67:121[Free Full Text]
  32. Laureys S, Goldman S, Phillips C, Van Bogaert P, Aerts J, Luxen A, Franck G, Maquet P. Impaired effective cortical connectivity in vegetative state: preliminary investigation using PET. Neuroimage 1999;9:377–82[Web of Science][Medline]
  33. Mason MF, Norton MI, Van Horn JD, Wegner DM, Grafton ST, Macrae CN. Wandering minds: the default network and stimulus-independent thought. Science 2007;315:393–5[Abstract/Free Full Text]
  34. Coull JT, Jones ME, Egan TD, Frith CD, Maze M. Attentional effects of noradrenaline vary with arousal level: selective activation of thalamic pulvinar in humans. Neuroimage 2004;22:315–22[Web of Science][Medline]
  35. Prielipp RC, Wall MH, Tobin JR, Groban L, Cannon MA, Fahey FH, Gage HD, Stump DA, James RL, Bennett J, Butterworth J. Dexmedetomidine-induced sedation in volunteers decreases regional and global cerebral blood flow. Anesth Analg 2002;95:1052–9[Abstract/Free Full Text]
  36. Fu CH, Reed LJ, Meyer JH, Kennedy S, Houle S, Eisfeld BS, Brown GM. Noradrenergic dysfunction in the prefrontal cortex in depression: an [15O] H2O PET study of the neuromodulatory effects of clonidine. Biol Psychiatry 2001;49:317–25[Web of Science][Medline]
  37. Giraud AL, Lorenzi C, Ashburner J, Wable J, Johnsrude I, Frackowiak R, Kleinschmidt A. Representation of the temporal envelope of sounds in the human brain. J Neurophysiol 2000;84:1588–98[Abstract/Free Full Text]
  38. Veselis RA, Feshchenko VA, Reinsel RA, Beattie B, Akhurst TJ. Propofol and thiopental do not interfere with regional cerebral blood flow response at sedative concentrations. Anesthesiology 2005;102:26–34[Web of Science][Medline]
  39. Plourde G, Belin P, Chartrand D, Fiset P, Backman SB, Xie G, Zatorre RJ. Cortical processing of complex auditory stimuli during alterations of consciousness with the general anesthetic propofol. Anesthesiology 2006;104:448–57[Web of Science][Medline]
  40. Campanella C, Salvini S, Casaldi S, Chiacchiararelli F, Serra A, Di Giacomo G. [Clonidine hallucinations: description of a clinical case]. Clin Ter 2000;151:45–7[Medline]
  41. Bodiou C, Bavoux F, Gavillon C, Krainik F. [Hallucinatory syndromes regressing after withdrawal of clonidine]. Therapie 1984;39:319–22[Web of Science][Medline]
  42. Brown MJ, Salmon D, Rendell M. Clonidine hallucinations. Ann Intern Med 1980;93:456–7[Abstract/Free Full Text]
  43. Shergill SS, Brammer MJ, Fukuda R, Williams SC, Murray RM, McGuire PK. Englagement of brain areas implicated in processing inner speech in people with auditory hallucinations. Br J Psychiatry 2003;182:525–31[Abstract/Free Full Text]
  44. Hennenlotter A, Schroeder U, Erhard P, Castrop F, Haslinger B, Stoecker D, Lange KW, Ceballos-Baumann AO. A common neural basis for receptive and expressive communication of pleasant facial affect. Neuroimage 2005;26:581–91[Web of Science][Medline]
  45. Kawase T, Yamaguchi K, Ogawa T, Suzuki K, Suzuki M, Itoh M, Kobayashi T, Fujii T. Recruitment of fusiform face area associated with listening to degraded speech sounds in auditory-visual speech perception: a PET study. Neurosci Lett 2005;382:254–8[Web of Science][Medline]
  46. Dietz NA, Jones KM, Gareau L, Zeffiro TA, Eden GF. Phonological decoding involves left posterior fusiform gyrus. Hum Brain Mapp 2005;26:81–93[Web of Science][Medline]
  47. Tan LH, Laird AR, Li K, Fox PT. Neuroanatomical correlates of phonological processing of Chinese characters and alphabetic words: a meta-analysis. Hum Brain Mapp 2005;25:83–91[Web of Science][Medline]
  48. Bitan T, Booth JR, Choy J, Burman DD, Gitelman DR, Mesulam MM. Shifts of effective connectivity within a language network during rhyming and spelling. J Neurosci 2005;25:5397–403[Abstract/Free Full Text]
  49. Hoenig K, Scheef L. Mediotemporal contributions to semantic processing: fMRI evidence from ambiguity processing during semantic context verification. Hippocampus 2005;15:597–609[Web of Science][Medline]
  50. Woodruff CC, Johnson JD, Uncapher MR, Rugg MD. Content-specificity of the neural correlates of recollection. Neuropsychologia 2005;43:1022–32[Web of Science][Medline]
  51. Lee HW, Caldwell JE, Dodson B, Talke P, Howley J. The effect of clonidine on cerebral blood flow velocity, carbon dioxide cerebral vasoreactivity, and response to increased arterial pressure in human volunteers. Anesthesiology 1997;87:553–8[Web of Science][Medline]
  52. Frisk-Holmberg M, Edlund PO, Paalzow L. Pharmacokinetics of clonidine and its relation to the hypotensive effect in patients. Br J Clin Pharmacol 1978;6:227–32[Web of Science][Medline]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a colleague
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via Web of Science (2)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bonhomme, V.
Right arrow Articles by Laureys, S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bonhomme, V.
Right arrow Articles by Laureys, S.
Related Collections
Right arrow Neuroanesthesia
Right arrow Preoperative Evaluation

Lippincott, Williams & Wilkins Anesthesia & Analgesia® is published for the International Anesthesia Research Society® by Lippincott Williams & Wilkins and Stanford University Libraries' HighWire Press®. Copyright 2008 by the International Anesthesia Research Society. Online ISSN: 1526-7598   Print ISSN: 0003-2999 HighWire Press