| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
 
|
|
|
||
|
|
Anesth Analg 2004;99:399-408 © 2004 International Anesthesia Research Society doi: 10.1213/01.ANE.0000131971.92180.DF
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
, |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Propofol has a greater amnesic effect than thiopental. In this study we tested whether different brain regions were affected by propofol and thiopental at similar drug effects. Changes in regional cerebral blood flow (rCBF) were identified by using SPM99 analysis of images obtained with positron emission tomography with 15O water. Ten right-handed male volunteers (age, 35 ± 10 yr; weight, 74.1 ± 7.5 kg; mean ± SD) were randomized to receive thiopental (n = 4) or propofol (n = 6) to target sedative and hypnotic concentrations with bispectral index (BIS) monitoring. Four positron emission tomography images were obtained during various tasks at baseline and with sedative and hypnotic effects. Two participants receiving propofol were unresponsive at sedative concentrations and were not included in the final analyses. Median serum concentrations were 1.2 and 2.7 µg/mL for sedative and hypnotic propofol effects, respectively. Similarly, thiopental concentrations were 4.8 and 10.6 µg/mL. BIS decreased similarly in both groups. The pattern of rCBF change was markedly different for propofol and thiopental. Propofol decreased rCBF in the anterior (right-sided during sedation) brain regions, whereas thiopental decreased rCBF primarily in the cerebellar and posterior brain regions. At similar levels of drug effect, propofol and thiopental affect different regions of the brain. These differences may help to identify the loci of action for the nonsedative effects of propofol, such as amnesia.
IMPLICATIONS: Despite very similar sedative and hypnotic effects, propofol and thiopental affect regional cerebral blood flow differently during drug effects. These differing patterns may provide clues as to how the drugs produce different behavioral effects, such as sedation and amnesia.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Drug-induced memory impairment is thought to result from both a sedative effect and a pure amnesic effect. Both occur at very similar drug concentrations, and these distinctions are difficult, but not impossible, to demonstrate behaviorally for propofol and thiopental. Specifically, with sufficiently large numbers of participants, propofol has additional effects on episodic memory compared with those of thiopental that are differentiable from the sedation caused by both drugs (1,2). However, there is considerable overlap in behavioral effects between drugs; thiopental demonstrates some memory effects even when corrected for sedation (1,3,4). Because both drugs are active at gamma-aminobutyric acid (GABA) receptors, differences in behavior may represent similar mechanisms of action with somewhat different expressions of effect or may represent actual differences in the mechanisms of action. Electrophysiologic markers, such as event-related potentials, provide some data that indicate different mechanisms of action, although variability in the data limits a conclusive inference (5).
Evidence of fundamentally different mechanisms underlying a behavior can derive from functional neuroimaging studies (FNI), which can localize the effect of interest with great spatial precision. Different patterns of activation during similar behaviors can provide strong evidence for fundamentally different mechanisms: one example is reading in dyslexic people. Such distinctions are most often made by using activation paradigms, wherein task-associated increases in the signal identify regions of interest. Regional cerebral blood flow (rCBF) changes are closely coupled with changes in neuronal activity, not only when neuronal activity is increased by increases in stimulus rate, but also by decreases induced by GABA agonists (6). The relation between increases in neuronal activity and rCBF responses is not affected in the presence of propofol or thiopental in sedative concentrations (7). Along these lines, two studies (8,9) have indicated that changes in rCBF localize the neuroanatomical targets of drug action in the brain.
Because thiopental and propofol produce very similar behavioral effects, the effect of these drugs on rCBF patterns was investigated. The hypothesis tested was that both drugs would produce fairly similar changes in the rCBF pattern, with possibly minor differences between these drugs. On the basis of behavioral differences, differences in the rCBF pattern would likely be difficult to ascertain in a typical number of participants in an FNI study. In this study, thiopental and propofol were administered to similar pharmacologic effects by using target concentrations based on previous experience. The neuroanatomical locations of drug effects were identified by using rCBF changes imaged with 15O water. Both drugs were expected to decrease global CBF (gCBF) as drug effects increased. Blood-flow patterns were analyzed with statistical parametric mapping (SPM) to identify regions where rCBF decreased to a greater extent than this global decrease ("decreases") and, conversely, where rCBF was resistant to this global decrease ("relative increases"). Because even small changes in PaCO2 can affect CBF (10), CBF patterns were analyzed with correction for PaCO2 values obtained from arterial blood samples taken during imaging.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This investigation was approved by the hospital IRB and the Radiation Safety Committee of Memorial Sloan-Kettering Cancer Center (New York, NY). Informed consent was obtained in writing before accrual of participants took place.
Ten healthy male volunteers aged 19 to 47 yr were recruited through newspaper advertisements and were paid for their participation. Participants were screened with telephone and subsequent in-person interviews. Exclusion criteria included current use of psychoactive medication; history of recreational drug abuse; head trauma resulting in loss of consciousness; neurologic, cardiovascular, or respiratory disease; claustrophobia; hypertension; peripheral vascular disease; hearing deficit; carpal tunnel syndrome; allergy to eggs (contraindication for propofol); a first-degree relative with schizophrenia; or a family history of acute intermittent porphyria (contraindication for thiopental).
The purpose of this study was to recreate a level of drug effect similar to those of our previous studies in which the memory and sedative effects of thiopental and propofol were closely investigated. Thus, target concentrations were chosen to produce similar sedative effects, on the basis of recent experience obtained with Stanpump (S. Shafer; http://anesthesia.stanford.edu/pkpd) (2). As measures of sedative effect, reaction time (RT), bispectral index (BIS), and measured serum concentrations of drug were obtained.
It is emphasized that this study examines the main effect of drug. Four positron emission tomography (PET) scans were obtained at each level of drug effect (no drug, sedative, and hypnotic target concentrations). At each level of drug effect, one PET scan was obtained during rest while two were obtained while words were presented through intraaural foam earphones, and one was obtained during tactile stimulation (vibration applied to right forearm). Task-related analyses will be reported separately. The four scanning conditions were administered in counterbalanced fashion across participants. Importantly, to allow the main effect of drug to be assessed, the sequence of conditions was the same for a given participant during drug administration. Regions affected by drug were identified in the statistical analysis as "relative decreases" or "relative increases" with respect to gCBF. As thiopental and propofol decrease gCBF, these relative statistical changes occur in the setting of decreased gCBF. Thus, relative increases are, in fact, regions of brain in which rCBF does not decrease as much as the change in gCBF.
At the orientation session, detailed information was given on study procedures, including radiation doses. Tests of handedness (Edinburgh Handedness Inventory, Oldfield 1971) and vocabulary (vocabulary subtest of the Wechsler Adult Intelligence Scale-Revised) were administered, followed by a brief physical examination. The session concluded with practice on the study tasks.
On the study day, participants arrived at approximately 8:00 AM, having had nothing by mouth since midnight. Participants were monitored with electrocardiograph (ECG) and pulse oximeter. Venous and radial arterial catheters were inserted before transfer to the PET suite. Saline with dextrose at approximately 100 mL/h was administered IV. Four baseline PET scans were obtained before drug administration, and these were repeated during drug infusion at sedative and hypnotic target concentrations. The volunteer’s total time in the PET scanner was approximately 3–4 h. After the completion of PET scanning, the arterial catheter was removed, and the volunteer was returned to the neuroanesthesia laboratory, where the IV was discontinued after the volunteer was given a light lunch. A delayed recognition test was administered shortly before discharge some hours later.
BIS (Version 1.3) was measured by using a standard clinical montage (Ziprep; Aspect Medical Systems, Natick, MA). Participants wore a 64-chanel electroencephalograph Quik-cap (NeuroMedical Supplies, El Paso, TX; data to be reported separately).
Volunteers were randomly assigned to receive propofol or thiopental administered by infusion with Stanpump software to target sedative (THP 4 µg/mL and propofol 1.2 µg/mL) and unresponsive (hypnotic target concentrations of 7–9 µg/mL for thiopental and 2.5–3 µg/mL for propofol) effects. Target concentrations were chosen on the basis of previous experience. The sedative concentration was chosen to provide maximal sedative and memory effects but allowed volunteers to remain responsive or arousable. The ability to achieve this state was variable, because participants easily fell asleep. Two of six participants receiving propofol were unresponsive at the sedative target concentration and were excluded from further analyses. The hypnotic concentration was more easily targeted, because a simple criterion of unresponsiveness to voice and light touch was tested. If participants responded, the target concentration was increased by 20% increments. After approximately 10 min, after predicted pseudoequilibration between serum and effect-site concentrations, PET scanning was resumed.
Arterial blood samples were obtained immediately after every scan for blood gas and drug assay. Propofol and thiopental concentrations in serum were determined by high-performance liquid chromatography with fluorescence detection, as described previously (11).
Volunteers were scanned with their eyes closed and wearing Earlink foam insert earphones (E*A*R Auditory Systems, Indianapolis, IN) for delivery of auditory stimuli during PET scanning (75 dB binaurally at 20 or 40 words per minute, started 20 s before PET acquisition). Participants were asked to pay attention and try to remember the words but did not make any response at the time. They had their heads resting on a foam finger mat for comfort. A U-shaped foam-lined plastic holder without a mask held the head stationary. During intervals between PET scans when no other tasks were being performed (approximately 5 min), participants were allowed to listen to music of their choice through headphones. Participants were instructed to remain awake during PET scanning for the baseline and sedation conditions.
A single 10-min transmission scan with a rotating rod of 68Ge/68Ga was performed before scanning commenced to correct for attenuation of the signal in its passage through bone and cerebral tissue. Four PET scans were obtained at each drug level, for a total of 12 scans. Every 12 min, 10 mCi of H215O IV was delivered at a constant rate over 20 s via an infusion pump. Scans were obtained on a GE Advance scanner (GE Medical Systems, Waukesha, WI) in the three-dimensional "septa out" mode. The resolution of the PET camera in this mode is approximately 5.2 mm in all dimensions. Three 30-s frames were obtained during each scan, corresponding to the highest rate of uptake of tracer into the brain. The three 30-s PET frames were analyzed as a single 90-s frame. The images were reconstructed by using filtered backprojection and standard clinical protocols and stored as "counts" images (counts of coincidence events expressed as nanocuries per cubic centimeter). Arterial blood was sampled continuously from the radial artery through a GE Fluid Radioactivity Quantifier during PET imaging to obtain a blood-activity time curve. The raw counts data from the PET image were transformed to quantitative CBF data by using an autoradiographic method after correction for difference in delay and dispersion of the measured blood-time activity curve. With this algorithm, the true arterial input function was determined as implemented in the GE Advance software. Quantitative blood-flow images were then transformed to ANALYZE format and subjected to SPM99 analysis.
After completion of scanning of the tactile stimulation task in each condition and before scanning commenced for the next task, participants performed a simple RT task to a series of 1000-Hz tones delivered every 1.5 s. They were instructed to respond to every tone by pressing a button as fast as possible. RT and the percentage of stimuli that received responses were recorded by a computer.
After the tactile scan, participants were asked to give numerical ratings on a standard Norris mood scale, previously described in detail (1). The physical and mental sedation scores from this scale were averaged together for analysis, because these have been previously shown to relate best to the sedative effects of these drugs (1). At the same time these subjective ratings were obtained, two observers (RAV and RAR) made independent ratings on the Observer Assessment of Alertness/Sedation (rated from 5 = alert to 1 = deep sleep). Ratings of the two observers were averaged for the composite measure of sedation level.
Performance data and participant-related information are presented throughout as median (range). For analyses of the behavioral and participant variables, nonparametric Mann-Whitney U-tests or Wilcoxon’s tests were used in view of the small number of participants in each group.
Quantitative blood-flow images were derived offline from the measured arterial input function and counts images by using an autoradiographic method as described above. These images were transformed to ANALYZE format for input into standard image-analysis software. Statistical analysis was performed by using SPM99 (Wellcome Institute of Cognitive Neurology, London; http://www.fil.ion.ucl.ac.uk/spm/spm99.html; implemented in Pro Matlab Version 6.5, Release 13; Mathworks, New York, NY) with an uncorrected threshold P value of 0.001 at a given voxel (corresponding to a T of 3.90).
Individual images were realigned to the last baseline scan. Individual PET images were co-registered to the individual’s structural T1-weighted magnetic resonance imaging (MRI) image before further analysis. These co-registered images were then transformed into standard image space consisting of an average MRI image derived from 152 normal structural MRI scans from the Montreal Neurologic Institute (Montreal, PQ, Canada) (the MNI coordinate system). A 16-mm Gaussian smoothing kernel was used to accommodate interpersonal variations in gyral anatomy and facilitate intersubject averaging with a resultant smoothness of 21.6 x 24.8 x 22.3 mm (x, y, z). Mean gCBF was normalized to 50 mL · 100 g–1 · min–1. Statistical analysis used a proportional scaling model, which allows a changing relationship between rCBF and gCBF depending on gCBF values. A significant regional effect in the SPM image was defined as a minimum volume of 82 voxels, which is the expected size of a random cluster. Arterial PaCO2 values obtained at the end of each scan were included in the SPM statistical design matrix to correct for any variability associated with this effect. SPM contrasts were constructed to identify regions of the brain demonstrating relative rCBF changes associated with the main effect of drug, as well as that of PaCO2 itself.
The traditional method of reporting the location of significant effects is to map locations of local maximum probability voxels to a standardized atlas, for example, the Talairach atlas. This works well for cognitive activations, for which discrete regions of the brain are affected. However, in most studies examining the effect of drugs, large regions of the brain are affected, and it is somewhat arbitrary to map only peak voxels. Thus, a method that summarizes all regions of the brain affected by drug is used here (corresponding to the colors overlaid on the structural images in Fig. 1). SPM99 provides x, y, and z coordinates by using the MNI coordinate system. The coordinate of the voxel with the maximal T value in a given brain region (cluster or "blob") is reported with this coordinate system (columns labeled x, y, and z in the tables). Separable regions of brain affected by drug (clusters) were mapped by using the MNI space utility program (Sergey Pakhomov; http://www.ihb.spb.ru/
pet_lab/MSU/MSUMain.html). This utility converts the MNI coordinates of significant voxels in each cluster (or discrete region of significant effect) into Talairach coordinates and labels these according to regions defined in this atlas. These are reported in the region and Brodmann’s areas (BA) columns by using the following criteria. If a significant portion of the cluster involved gray matter (>25% of all voxels), the BAs and corresponding regions were reported in decreasing order of the amount of activation present in the region. In general, 5%–20% of the activation was present in a reported BA. However, some BAs are relatively small, and even though it contains a small percentage of total activation, a large portion of the BA may be affected (e.g., in Table 1, 9% of the 5726-voxel, first cluster under "Hypnosis, Propofol Hypn rCBF decrease" was present in BA 39, but 70% of this BA was affected by this concentration of propofol). In this case, the BA was reported as being of interest. If more than 75% of the voxels were in white matter, the affected region was reported as "white matter" only.
|
|
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
One propofol and one thiopental participant dozed off during word presentation but were responsive to voice command immediately after the end of the scan. Two participants in the propofol group were unresponsive during the sedation condition and were excluded from further analyses, including the hypnotic effect analysis. By design, all participants were unresponsive at the hypnotic dose. Transient myoclonus was noted in one participant who received propofol in the sedation phase and in two other participants (one propofol and one thiopental) during the hypnotic stage. All participants responded to verbal stimulation within a few minutes after discontinuation of the infusion.
The 8 volunteers included in the analyses ranged in age from 21 to 47 yr (mean ± SD: 36 ± 10 yr) and weighed between 60.5 and 85 kg (74.1 ± 7.5 kg). All volunteers had received at least some college education (four were college graduates, two of whom had some postgraduate education), were right-handed, and were native English speakers with normal hearing and vocabulary ability (mean vocabulary percentile, 78.2 ± 21.7). All volunteers reported sleeping normally the night before, although one slept only 5 h. Participants receiving propofol or thiopental did not differ on any variable. The 2 participants receiving propofol who were unresponsive at the sedative target concentration were aged 19 and 30 yr with body mass indexes of 23.8 and 26.8 kg/m2. During sedation, their BIS values were 65.9 and 80.3.
Both propofol and thiopental produced significant regional effects on CBF. These are reported in Tables 1 and 2 and are displayed in Figure 1. Sedative drugs produce global decreases in CBF, but this effect was nullified in the SPM analysis with global normalization. Thus, rCBF decreases represent regional changes in CBF beyond any global decrease. Conversely, rCBF increases likely represent regions of the brain that are relatively unaffected by drug effects rather than representing true increases in rCBF and are referred to as "relative increases" to emphasize this relationship (12).
|
The locations of rCBF changes are quite different for both drugs. The rCBF decreases are visualized in Figure 1, where propofol effects are displayed in warm colors and thiopental effects are displayed in cool colors. At sedative concentrations, affected regions are usually mutually exclusive. At hypnotic concentrations, overlap increases; these regions are identified by purplish colors.
At sedative effect, propofol decreases rCBF primarily in the right-sided anterior brain, whereas thiopental decreases rCBF primarily in the left-sided cerebellum. At this level of drug effect, both drugs have an asymmetrical effect. At hypnotic concentrations, effects become more symmetrical. However, the differing distribution of anterior versus posterior propofol versus thiopental effects, respectively, is still largely present. At hypnotic effect, both drugs decrease rCBF in posterior cortical regions. Interestingly, at hypnotic effect propofol decreases rCBF in the thalamus, whereas thiopental does not.
Approximately half the voxels demonstrating rCBF decrease map to cortical regions, and the other half map to white matter. However, most voxels demonstrating relative rCBF increases map to white matter (60%–80% versus 19% mapping to gray matter).1 Interestingly, propofol seems to have relatively little effect on rCBF in the medial temporal lobe regions and cerebellum. Conversely, thiopental demonstrates a relative rCBF increase in the medial frontal and rectal/orbital gyri, some of which demonstrate rCBF decreases with propofol.
The median (range) sedative and hypnotic concentrations were 1.2 µg/mL (1.1–1.4 µg/mL) and 2.7 µg/mL (2.4–2.7 µg/mL), respectively, for propofol and 4.8 µg/mL (3.5–5.7 µg/mL) and 10.6 µg/mL (8.5–14.1 µg/mL), respectively, for thiopental. For all participants combined, regardless of the drug administered, the median PaCO2 was 45.5 mm Hg (range, 39–51 mm Hg) during baseline and was unchanged during sedation (median, 44.9 mm Hg; range, 42.5–50 mm Hg). At hypnotic concentrations, PaCO2 increased significantly to 50.1 mm Hg (44.7–53 mm Hg) (P = 0.002 by nonparametric Friedman analysis of variance). There were no differences between the two drugs at any condition (Table 3).
|
BIS decreased significantly during both sedative (P = 0.025) and hypnotic (P = 0.012) drug concentrations (Table 3). Thiopental participants had nonsignificantly lower mean BIS values at baseline than propofol participants (P < 0.09). There were no significant differences between groups at sedative or hypnotic effects, although during hypnosis one volunteer with a relatively high BIS of 83 caused the BIS values in the thiopental group to be nonsignificantly higher than in the propofol group (P < 0.09).
There was a significant increase in RT during sedation for all participants taken together (P = 0.012 by Wilcoxon test), as shown in Table 3. Overall, there were no differences between drug groups for RT.
Participants scored themselves as only slightly less alert (P < 0.06) during sedative doses of propofol or thiopental than at baseline. The increase in sedative ratings from baseline to sedation was not significant for either propofol or thiopental (Table 3). Ratings were not obtained at hypnotic doses because participants were unresponsive. Baseline Observer Assessment of Alertness/Sedation ratings averaged close to the maximum score of 5, were lower during sedation (P < 0.03), and decreased further during unresponsiveness (P < 0.02; Table 3). At hypnotic doses, all four propofol participants were rated as unresponsive, whereas participants receiving thiopental (data were available for three patients) might have been more responsive (P < 0.08 by Mann-Whitney U-test). Across conditions, the 2 raters correlated highly with each other (0.93 for propofol and 0.77 for thiopental; P < 0.01 in both cases, two-tailed, by nonparametric Spearman rank correlation).
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Both propofol and thiopental cause significant decreases in rCBF beyond any changes in gCBF. These changes become more widespread as the drug effect increases. The striking finding is that, despite very similar behavioral states, these drugs affect rCBF in distinctly different regions of the brain. These patterns were quite different even in this small number of volunteers. No difference in behavioral effect, as far as memory is concerned, would be present in such a small number of participants. Differences in memory at this level of drug effect have been demonstrated in larger numbers of participants. The striking difference in rCBF effects is even more intriguing because both drugs act at GABA receptors. In fact, some regions of the brain that are resistant to drug effect (relative rCBF increases) with one drug are regions of the brain that demonstrate rCBF decreases with the other drug, such as the medial frontal gyri.
There is good circumstantial evidence that neuroanatomical structures affected by anesthetic drugs are identified by depressions in rCBF or blood oxygen level-dependent (BOLD) imaging signals. In studies where no drug is administered, changes in rCBF represent changes in local neuronal activity, as mediated by changes in neuronal metabolism and blood flow. In many ways, the BOLD signal measured with functional MRI is analogous to rCBF change with neuronal activation, and both are included in the term FNI. The neuroanatomical location and structures involved with a particular behavior, such as moving one’s finger, can be imaged with great precision with FNI. By extension, more complicated cognitive behaviors can be associated with given neural structures and networks. However, information and its transmission in the brain are represented by neuronal spike activity, which is only indirectly measured by FNI. The local hemodynamic changes measured by FNI are more closely related to local synaptic activity (13). Usually spike activity and synaptic activity are closely related, particularly when excitatory synapses predominate in a given region (14,15). However, this relationship may become less tightly coupled, particularly when inhibitory synaptic activity predominates (16,17). There may be regional differences in this relationship, and dissociations have been found in the cerebellum, for example. These relationships still require further investigation, particularly situations in which drugs with systemic physiologic effects are being given. Despite these considerations, on the whole, when sedative/hypnotic drugs are administered, decreases in the FNI signal most likely indicate local decreases in synaptic activity associated with drug effects.
A related issue is that of the coupling of metabolic changes with blood-flow changes, which theoretically might be uncoupled by anesthetics. These could produce a heterogenous "pharmacologic" vasodilation unrelated to underlying neuronal activity or metabolism. Thus, FNI signal decreases in this setting would not indicate decreased neuronal synaptic activity as decoupling occurs. Increasing evidence indicates that coupling is most likely preserved when clinical doses of anesthetics or sedatives are used. For example, during sedation with lorazepam or scopolamine resulting in memory impairment, the BOLD signal from the hippocampus is inhibited without change in the visual cortex (18). The relationship among rCBF, regional cerebral blood volume (CBV), and oxygen consumption is largely maintained with anesthetics that cause global decreases (19) or increases (20) in CBF. A number of studies indicate that changes in rCBF during drug administration do, in fact, correspond to changes in neuronal activity. Specific regional changes in CBF in the thalamus correspond closely to changes in consciousness, beyond the global decreases in CBF with propofol (21). Alternatively, evidence of neuronal activation in the auditory and tactile cortices during heavy propofol sedation/light anesthesia is present, even during unresponsiveness (7,21).
Whether changes in glucose metabolism or blood flow are better at localizing the site of anesthetic action on the brain is an open question. Changes in glucose metabolism may exclude any considerations about coupling, because blood flow is not measured and may or may not be spatially linked to receptor density (22). Changes in rCBF may be distant from the drug/receptor interaction (23). This likely occurs when synaptic inhibition results from a drug/receptor interaction at a distant location. However, changes in rCBF that are indicative of synaptic effects may be functionally more relevant than the location of changes in the regional cerebral metabolic rate for glucose or receptor binding, because synaptic activity may relate more closely to the function of neurons in networks than to the distribution of receptor density.
A number of inferences may be postulated on the basis of the markedly different patterns of rCBF changes induced by propofol and thiopental. It could be that differences in rCBF patterns will help to identify neuroanatomical regions that mediate differential effects of drugs. However, it is somewhat surprising that such different patterns in rCBF effect were seen in such a small number of participants; this is contrary to the similarity of behavioral effects. Thus, the relation between rCBF change in behavior is likely to be more complex. For example, these rCBF patterns may represent downstream effects of the drug at a yet-to-be-identified locus. Even though drugs may act at a similar locus, rCBF markers of this effect may be different on the basis of different properties of transmission in the brain modulated by drug effect. An analogous effect occurs with electroencephalogram rhythms under the influence of drug, exemplified by resting 
rhythms (24). The rhythm likely represents the filtering properties of local cortical neuronal ensembles on a driving input from deeper cerebral structures. The properties of these ensemble filters are changed differently by different sedative drugs, resulting in differentiable effects on cortically expressed rhythms. Yet another possibility is that numerous interrelated regions are involved in the maintenance of normal memory function or attention, the inhibition of which is manifested as amnesia or sedation, respectively. Inhibition of a subset of these regions will result in failure of the behavior, with a graded response being produced by more and more regions being affected. Thus, thiopental and propofol both produce sedation but by inhibition of different brain regions, all of which may be important in the maintenance of attention. Another, less likely, possibility is that drug-induced rCBF changes bear no relationship to drug effects on behavior. A likely explanation in this case is that different drugs have different patterns of effect on cerebral vasculature. In fact, there is evidence that changes in PaCO2, for example, result in differential changes in rCBF (10). Although the regional effects of various drugs on cerebral circulation have been previously described (25), the interaction of rCBF and neuronal activity in the presence of drug at the level of detail available in current neuroimaging instruments has not been well studied. On a microscopic level, anesthetics diminish signal intensity, but activation in the signal is still closely coupled with neuronal activation (26).
Despite some difficulty in differentiating differences in drug effects by behavioral measures (1), these rCBF findings do indicate that these drugs could be affecting the brain by differing mechanisms. We suggest that the most parsimonious explanation is one in which regions affected by propofol but not thiopental (or vice versa) represent those whose inhibition results in drug-induced amnesia and sedation. We postulate that the amnesic effect of propofol is mediated primarily in the anterior brain, with specific depressant effects in the ventrolateral prefrontal cortical regions and the insula, represented by rCBF decreases in BAs 44, 45, and 13. These prefrontal regions have been identified not only with working memory processes, but also with long-term memory processes, including encoding, retrieval and recognition, and emotively mediated executive processes (27). The insula seems to act as an interface between the affective limbic system and higher-order executive processes mediated in the prefrontal cortex. The sedative effects of thiopental may be mediated by the cerebellum. At larger concentrations, both drugs begin to affect similar cerebral structures. Because propofol’s amnesic effects occur at smaller concentrations than its sedative effects, these posterior cortical structures may represent neuroanatomical locations that also mediate sedation.
Because of the small sample size in this study, any behavioral measures will be somewhat unreliable. Volunteers had differentiable memory effects in previous studies at the serum concentrations present during the sedation condition in this study (1,3). Previously determined values of the dose which produces sedation in 50% of individuals for propofol and thiopental in a similar population were 1.9 and 4.8 µg/mL, respectively (1). These are close to actual serum concentrations measured at the sedative concentration in this study. By subjective and RT measures, a discernible and equivalent degree of sedation was present. The degree of sedation and hypnosis as measured by BIS was also equivalent between groups. Thus, as near as can be determined, propofol and thiopental were administered to very similar behavioral and pharmacologic levels in this study.
An important confound is that of changes in PaCO2, which is almost inevitable in this type of study unless ventilation is controlled by mechanical ventilation of an unresponsive participant, for example. Changes in PaCO2 associated with sedation will potentially cause regional variation in signal response (10). In this study, we explicitly included PaCO2 measures in the statistical model used to detect rCBF changes. The rCBF changes imaged in different drug conditions exclude any effect of PaCO2 on rCBF. An analysis of rCBF change in relation to PaCO2 reveals a significant effect from PaCO2 change itself only in a peripheral region of the left temporal lobe (MNI coordinates –30, 4, –42; T = 5.19; cluster size, 677 voxels; uncus, superior temporal gyrus, BA 38), which is distant from most regions of the brain that demonstrate drug effects.
It is interesting that the thalamus is affected by propofol, but not by thiopental, despite similar behavioral states of unresponsiveness. It has been postulated that loss of responsiveness is mediated by a failure of transmission of sensory information through the thalamus (thalamocortical switch) (21,28). We know of no other study in humans that has examined rCBF changes with thiopental. The findings of this study, obtained in a small number of participants, are preliminary, and with more participants or at a greater drug effect, thiopental may also demonstrate decreases in thalamic rCBF.
It is interesting that regions of the brain that exhibit a relative resistance to drug effects are primarily located in white matter. This observation was derived by analysis of neuroimaging data by using the total volume of significant effect, rather than by mapping a local maximum. White matter, comprising the myelinated "cabling" of the brain, has been poorly studied with FNI techniques. There is some evidence that metabolic changes in white matter occur in direct relationship to cortical activation or stimulation (29). It may be that the depressant effects of anesthetic drugs are focused on synaptic activity in cortical structures rather than on general metabolic activity used for maintenance of suitable ionic gradients that are most evident in white matter.
In conclusion, the dramatically different pattern of rCBF changes of propofol from thiopental with similar behavioral effects indicates different mechanisms of action of these drugs. By further delineation of these effects, specific regions of the brain may be identified whose inhibition results in cognitive changes such as sedation and drug-induced amnesia.
|
Acknowledgments |
|---|
Funded by National Institutes of Health Grant RO1-58782.
We thank Avin S. Lalmansingh, BA, and Olivia Squire, RN, for assistance in conducting the PET studies. We thank Aspect Medical Systems for loaning the equipment used to collect BIS data. We especially thank our volunteer research participants for their patience and cooperation.
|
Footnotes |
|---|
Presented at the 77th International Anesthesia Research Society Clinical and Scientific Congress, New Orleans, LA, March 21–25, 2003.
1 A portion of voxels are "unidentified," meaning they are not located in gray matter, white matter, or cerebral spinal fluid (e.g., those in cerebellum). A portion of voxels will be located outside the standardized Talairach atlas, simply on the basis of deformation of spatially filtered data into a standard image space. 
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
- Veselis RA, Reinsel RA, Feshchenko VA, Wronski M. The comparative amnestic effects of midazolam, propofol, thiopental, and fentanyl at equisedative concentrations. Anesthesiology 1997; 87: 749–64.[Web of Science][Medline]
- Veselis RA, Reinsel RA, Feshchenko VA, Parsons J. Amnesia from propofol is characterized by information loss during a critical time period [abstract]. Anesthesiology 2003; 99: A770.
- Reinsel R, Veselis R, Wronski M, et al. Memory impairment during conscious sedation: a comparison of midazolam, propofol and thiopental. In: Sebel PS, Bonke B, Winograd E, eds. Memory and awareness in anesthesia. Englewood, NJ: Prentice-Hall, 1993: 127–40.
- Osborn AG, Bunker JP, Cooper LM, et al. Effects of thiopental sedation on learning and memory. Science 1967; 157: 574–6.
[Abstract/Free Full Text] - Veselis RA, Reinsel RA, Feshchenko VA. Drug-induced amnesia is a separate phenomenon from sedation: electrophysiologic evidence. Anesthesiology 2001; 95: 896–907.[Web of Science][Medline]
- Jueptner M, Weiller C. Review: does measurement of regional cerebral blood flow reflect synaptic activity? Implications for PET and fMRI. Neuroimage 1995; 2: 148–56.[Web of Science][Medline]
- Veselis RA, Reinsel RA, Feshchenko VA, et al. Auditory rCBF covariation with word rate during drug-induced sedation and unresponsiveness: an H2O15 study [abstract]. Brain Cogn 2004; 54: 142–4.[Web of Science][Medline]
- Chen YC, Mandeville JB, Nguyen TV, et al. Improved mapping of pharmacologically induced neuronal activation using the IRON technique with superparamagnetic blood pool agents. J Magn Reson Imaging 2001; 14: 517–24.[Web of Science][Medline]
- Jenkins BG, Chen IYI, Brownell A, et al. Mapping dopamine receptors in the rodent and primate brain with IRON MRI: comparisons with dopamine receptor sub-type distributions [abstract]. Neuroimage 2001; 14: 985.
- Ito H, Yokoyama I, Iida H, et al. Regional differences in cerebral vascular response to PaCO2 changes in humans measured by positron emission tomography. J Cereb Blood Flow Metab 2000; 20: 1264–70.[Web of Science][Medline]
- Veselis RA, Glass P, Dnistrian A, Reinsel R. Performance of computer-assisted continuous infusion at low concentrations of intravenous sedatives. Anesth Analg 1997; 84: 1049–57.[Abstract]
- Veselis RA, Reinsel RA, Beattie BJ, et al. 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]
- Logothetis NK, Pauls J, Augath M, et al. Neurophysiological investigation of the basis of the fMRI signal. Nature 2001; 412: 150–7.[Medline]
- Heeger DJ, Huk AC, Geisler WS, Albrecht DG. Spikes versus BOLD: what does neuroimaging tell us about neuronal activity? Nat Neurosci 2000; 3: 631–3.[Web of Science][Medline]
- Rees G, Friston K, Koch C. A direct quantitative relationship between the functional properties of human and macaque V5. Nat Neurosci 2000; 3: 716–23.[Web of Science][Medline]
- Logothetis NK. The underpinnings of the BOLD functional magnetic resonance imaging signal. J Neurosci 2003; 23: 3963–71.
[Free Full Text] - Almeida R, Stetter M. Modeling the link between functional imaging and neuronal activity: synaptic metabolic demand and spike rates. Neuroimage 2002; 17: 1065–79.[Web of Science][Medline]
- Sperling R, Greve D, Dale A, et al. Functional MRI detection of pharmacologically induced memory impairment. Proc Natl Acad Sci U S A 2002; 99: 455–60.
[Abstract/Free Full Text] - Kaisti KK, Metsahonkala L, Teras M, et al. Effects of surgical levels of propofol and sevoflurane anesthesia on cerebral blood flow in healthy subjects studied with positron emission tomography. Anesthesiology 2002; 96: 1358–70.[Web of Science][Medline]
- Langsjo JW, Kaisti KK, Aalto S, et al. Effects of subanesthetic doses of ketamine on regional cerebral blood flow, oxygen consumption, and blood volume in humans. Anesthesiology 2003; 99: 614–23.[Web of Science][Medline]
- Bonhomme V, Fiset P, Meuret P, et al. Propofol anesthesia and cerebral blood flow changes elicited by vibrotactile stimulation: a positron emission tomography study. J Neurophysiol 2001; 85: 1299–308.
[Abstract/Free Full Text] - Menon DK. Mapping the anatomy of unconsciousness: imaging anaesthetic action in the brain. Br J Anaesth 2001; 86: 607–10.
[Free Full Text] - Petrovic P, Tolle TR, Ingvar M, et al. Opposite opioid effect on rCBF in regions with high opioid receptor binding [abstract; http://208.164.121.55/hbm2003/abstract/abstract304.htm]. Neuroimage 2003(Suppl 1);19.
- Feshchenko VA, Veselis RA, Reinsel RA. Comparison of the EEG effects of midazolam, thiopental, and propofol: the role of underlying oscillatory systems. Neuropsychobiology 1997; 35: 211–20.[Web of Science][Medline]
- Madsen JB, Cold GE. General considerations concerning the effects of anaesthetics on cerebral blood flow and metabolism: experimental and clinical studies—the effects of anaesthetics upon cerebral circulation and metabolism. Vienna: Springer Verlag, 1990: 2–3.
- Shtoyerman E, Arieli A, Slovin H, et al. Long-term optical imaging and spectroscopy reveal mechanisms underlying the intrinsic signal and stability of cortical maps in V1 of behaving monkeys. J Neurosci 2000; 20: 8111–21.
[Abstract/Free Full Text] - Ranganath C, Johnson MK, D’Esposito M. Prefrontal activity associated with working memory and episodic long-term memory. Neuropsychologia 2003; 41: 378–89.[Web of Science][Medline]
- 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]
- Weber B, Fouad K, Burger C, Buck A. White matter glucose metabolism during intracortical electrostimulation: a quantitative [(18)F]fluorodeoxyglucose autoradiography study in the rat. Neuroimage 2002; 16: 993–8.[Web of Science][Medline]
This article has been cited by other articles:
|
|
 |
|
  M. T. Alkire, A. G. Hudetz, and G. Tononi Consciousness and Anesthesia Science, November 7, 2008; 322(5903): 876 - 880. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U. Freo, M. Dam, and C. Ori The Time-Dependent Effects of Midazolam on Regional Cerebral Glucose Metabolism in Rats Anesth. Analg., May 1, 2008; 106(5): 1516 - 1523. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. G. Bovill Anesthetic Pharmacology: Reflections of a Section Editor Anesth. Analg., November 1, 2007; 105(5): 1186 - 1190. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
