The issue of how and where general anaesthetics act in the brain continues to intrigue and occupy researchers. Studies that address these issues can be focused at different levels. One reductionist approach is to investigate anaesthetic interactions with specific receptors or other molecular targets in a variety of models.16 While such studies are important, they are difficult to interpret in the context of functional neuroanatomical models of consciousness.7 8 An alternative approach is to consider the differential effects of anaesthesia on different brain areas, with a view to identifying specific brain regions that are important for anaesthesia (and by inference, for the generation of consciousness).
Such an assessment of spatial variations in general anaesthetic effects in the brain is not easy. While clinical measurement of anaesthetic effect has substantially depended on measuring spontaneous or evoked electrical responses,7 current implementations cannot provide the tomographic visualization of subcortical physiology required to understand sites of anaesthetic action.8 9 Magneto-encephalography (MEG) can provide detailed information regarding foci of brain activation,10 but while the technique has been applied to pre-surgical planning,11 there are no reports of MEG studies of anaesthetic action.
This inability to map primary brain function (i.e. electrical activity) has led researchers to image secondary physiological phenomena associated with neuronal excitation. As neuronal activity is closely coupled to glucose metabolism (and in many circumstances to blood flow), imaging the effect of anaesthetic agents on cerebral blood flow (CBF) and cerebral metabolic rates for glucose (CMRgluc) may provide surrogate measures of regional effects. Indeed, these indices have been investigated in several experimental studies of anaesthetic action in animal models using autoradiography with [14C]iodoantipyrine and [14C]deoxyglucose, respectively.1217 More recent experimental studies have used the newer technique of functional magnetic resonance imaging (fMRI) to map changes in regional CBF associated with anaesthetic effects.18 Similar studies are now possible in humans with techniques such as positron emission tomography (PET) and fMRI.
The most common implementation of fMRI in this context uses blood oxygen level dependent (BOLD) contrast,19 20 which depends on coupled increases in regional CBF associated with neuronal activation. It is now known that such CBF recruitment is associated with a reduction in proportional oxygen extraction, and a relative increase in regional oxygen saturation. The use of appropriate image acquisition protocols or sequences translates these changes in regional oxygen saturation into areas of high signal intensity on subtraction images. These changes can be submitted to statistical analysis to identify areas of significant blood flow change, and inferences drawn about neural activation. The microcirculatory changes responsible for such fMRI contrast are crucially dependent on the maintenance of normal flow-metabolism coupling. While flow-metabolism coupling is retained to a substantial extent with many anaesthetics,21 the relative difficulty of providing anaesthesia in MR environments has limited the application of fMRI to studies of anaesthetic action in humans. However, one recent fMRI study in human volunteers suggested that isoflurane impairs thalamo-cortical transmission of sensory information,22 in keeping with previous electrophysiological studies in animal models.8
Positron emission tomography studies the distribution and kinetics of molecules that incorporate positron-emitting isotopes2325 such as H215O, which is used to study blood flow, and 18fluoro-deoxyglucose (18FDG), which provides images of cerebral glucose uptake. Such studies provide maps of cerebral physiology and metabolism with sub-centimetre spatial resolution and a sufficiently small radiation burden (typically less than 5 mSv in the context of an annual background radiation in East Anglia of about 2 mSv) that allows application of the technique to healthy volunteers. While concomitant sampling of radioactivity in arterial blood can provide quantitative results, it is far more common in volunteer studies to acquire non-quantitative images. This approach has been widely used by neuropsychologists.26 Blood flow maps are acquired during a selected control task and during a test task, which imposes a small and well-defined additional cognitive burden on the brain. Differences in CBF patterns observed between the two tasks can then be attributed to neuronal activation and the synaptic activity required to service the additional cognitive burden of the test task27. While this process of cognitive subtraction has occasionally been criticized,28 29 it (or one of its variants) still forms the basis of most functional imaging studies. This technique can be extended to studies of drug effects on the brain, where the cognitive burden is invariant, but images are acquired before and after a drug is administered. As with conventional functional imaging, subtraction images can be submitted to careful statistical analysis to identify significant increases or decreases in CBF or CMRgluc associated with administration of a drug.
Several research groups have attempted to use PET to identify sites of anaesthetic action in the human brain,3034 and have provided interesting, though not always concordant results. Alkire and colleagues used 18FDG PET to show that halothane30 and isoflurane31 resulted in global reductions in cerebral metabolism, with relatively prominent effects on the basal forebrain structures, thalamus, limbic system, and cerebellum. In contrast, propofol32 produced more metabolic depression in the cortex than in subcortical structures. However, these data are at odds with a more recent study that used H215O PET to study changes in CBF during graded propofol anaesthesia.33 These authors showed global reductions in CBF with relatively prominent effects in the medial thalamus, in addition to specific cortical areas. They also demonstrated a close correlation between midbrain and thalamic blood flow, suggesting concordant changes in arousal systems. Interestingly, in another study, high dose midazolam34 resulted in CBF reductions in multiple cortical areas, but also selectively reduced thalamic CBF. Specific cortical areas that were commonly affected by both agents across three studies included the angular gyrus, anterior cingulate area, and parietal and temporal association cortices, areas that are commonly implicated in arousal and information processing. While it is tempting to suggest that these data reveal the anatomical sites at which propofol exerts its anaesthetic effects, it is also possible that these changes are a consequence of the anaesthetic state produced by propofol acting at a more strategic and focused site.
The functional imaging studies described thus far address the where of anaesthetic action in neuroanatomical terms, but it may be possible to provide additional specificity in terms of the receptor systems involved. Data regarding the regional distribution of receptor subtypes are available from post-mortem and PET ligand studies. Synthesis of such data with changes in regional CBF or CMRgluc would allow us to correlate the distribution of a putative site of anaesthetic action with the documented distribution of regional pharmacodynamic effects.
The paper by Alkire and colleagues35 in this issue of the journal attempts to do this by using PET data that has, in large part, been previously published by these authors. They relate the cerebral metabolic effects of isoflurane and propofol in human subjects (measured using 18FDG PET)31 32 to the regional distribution of selected neurotransmitter systems in the brain (using literature data from post-mortem immunohistochemistry36 37 to define receptor densities). In principle, the approach is a valid one, and has produced interesting results, with dissociation demonstrated between patterns of metabolic suppression seen with propofol and isoflurane. The authors also demonstrate a correlation of regional metabolic suppression by propofol with historical data on regional [3H]diazepam and [3H]flunitrazepam binding. Unexpectedly, isoflurane induced reductions in CMRgluc did not correlate with benzodiazepine binding, but showed an inverse correlation with muscarinic receptor density. These data led the authors to conclude that the most logical interpretation for this effect is to suggest that the diazepam binding site on the GABAA complex is likely to be strongly regionally co-localized with the GABAergic site that mediates propofols in vivo effects on brain metabolism. They suggest that the absence of such a correlation with isoflurane may be because of a variety of factors, including the possibility that it acts on several receptor systems, and to use common parlance, is a dirtier agent. They suggest that the inverse correlation seen with muscarinic receptor binding can be accounted for by the well documented arousal effects of this receptor system, which would antagonize the metabolic suppression produced by isoflurane in proportion to its local density.
While these data are interesting, several caveats need to be voiced regarding the details of the methodology used for functional imaging studies of anaesthetic action in general, and the current paper in particular. These techniques are based on the assumption that coupled increases in blood flow and glucose metabolism can provide a marker of regional neuronal activation. While this may be true for physiological activation, its relevance to pharmacological deactivation remains unclear. In particular, caution needs to be exercised regarding the implicit but central assumption that a reduction in regional brain metabolism by a drug is necessarily the consequence of a modulation of synaptic activity by the agent in that region. While this may well be the case, this is not always so, as suppression of strategic neuronal pathways can clearly affect metabolic activity in their projection areas. For example, marked suppression of thalamic transmission of somatosensory inputs can decrease cortical activity in the somatosensory cortex. Indeed, Angel and colleagues8 have shown that such thalamic activity is an important site of action for several anaesthetic agents, including those thought to act primarily at the GABAA complex. These electrophysiological data are concordant with the fMRI data from Antognini22 who demonstrated lack of cortical somatosensory activation following anaesthesia despite preserved thalamic activation, implying arrest of inputs at a thalamic gate.22 38 In this situation, the visible deactivation is in the cortex, but the site of anaesthetic action is the thalamus.
The direct vascular effects of anaesthetics present us with another confounder in such studies, especially when the paradigm involves higher doses of halogenated volatile agents.39 The varying degrees of direct cerebral vasodilation observed with these agents may modulate the CBF responses to activation and deactivation. However, available evidence suggests that even drugs that alter neurovascular coupling do not alter changes in cerebral glucose metabolism, other than by direct neuronal effects.40
Studies also need to take account of the variability in functional neuroanatomy between subjects. While these considerations may not be relevant to studies addressing gross patterns of activation and deactivation, they may be important when testing the effects of low or residual levels of anaesthesia in functional imaging paradigms that assess effects on complex cognitive tasks. In these circumstances, functional images should be coregistered to each individual subjects anatomy (using a high resolution MRI), and image averaging undertaken after normalizing these images to a generic template. Ideally, regions of interest that define neuroanatomical structures should be identified on such high-resolution images.
Finally, the source of data for receptor densities in comparisons of the kind used by Alkire and colleagues35 remains a difficult issue. There are problems using in vitro data on [3H]diazepam binding36 as a basis for assessing regional GABA receptor density. Diazepam binds to both neuronal (GABA/Cl channel associated) and mitochondrial benzodiazepine receptors (MBRs).41 MBRs are typically expressed by inflammatory cells, including activated microglia and blood derived macrophages. Their use of flunitrazepam binding to corroborate these data substantially addresses this criticism, as this is a selective neuronal benzodiazepine ligand.37 However, regardless of ligand used, post-mortem data cannot account for inter-individual variability, and an ideal study would have compared regional benzodiazepine receptor density in volunteers with regional reductions in 18FDG uptake induced by anaesthetic agents in the same subjects. This is a technically feasible exercise, since [11C]flumazenil is a well-characterized PET ligand in humans.42
Even if we accept that post-mortem benzodiazepine binding provides an acceptable approximation of central benzodiazepine receptor distribution, caution is required in equating this to GABAA receptor distribution or to the regional metabolic effects of potentiating GABAergic receptors. Several papers43 44 (including that by Braestrup and colleagues36) make the point that regional [3H]diazepam binding does not necessarily correlate with GABA receptor distribution (though there is a correlation with bicuculline binding sites). More importantly, administration of GABA agonists such as muscimol and THIP (and diazepam) to animals does not produce changes in glucose metabolism in patterns that correlate to receptor density.43 44
Despite these reservations, the authors findings are intriguing. The relative selectivity of propofol induced cortical metabolic suppression (in contrast to that produced by isoflurane, which was more global) may well be caused by variations in the regional density of receptor systems. However, the data provided do not prove this (and to be fair, the authors do not claim it). What the authors have done is to provide a clear experimental basis for an interesting and testable hypothesis. Using modern PET cameras, it should be possible to obtain 18FDG studies at baseline and with anaesthetics, [11C]flumazenil PET images and anatomical MR, for coregistration in a cohort of volunteers with an acceptable radiation burden (6 mSv) to participants. Indeed, modern PET ligand technology already offers, or promises to provide, access to a wide variety of receptors45 (including opioid4648 and cholinergic4951) that would be of interest in this context. Future results from such studies promise to be of great interest.
D. K. Menon
Professor of Anaesthesia, University of Cambridge
Director, Neurocritical Care Unit
Addenbrookes Hospital
Cambridge
UK
References
1 Van Swinderen B, Shook DR, Ebert RH, et al. Quantitative trait loci controlling halothane sensitivity in Caenorhabditis elegans. Proc Natl Acad Sci USA 1997; 94: 82327
2 Krasowski MD, Koltchine VV, Rick CE, et al. Propofol and other intravenous anesthetics have sites of action on the -aminobutyric acid type A receptor distinct from that for isoflurane. Mol Pharmacol 1998; 53: 5308
3 Franks NP, Lieb WR. Selectivity of general anesthetics: a new dimension. Nat Med 1997; 3: 3778
4 Lees G. Molecular mechanisms of anaesthesia: light at the end of the channel? Br J Anaesth 1998; 81: 4913
5 Franks NP, Lieb WR. Which molecular targets are most relevant to general anaesthesia? Toxicol Lett 1998; 100-101: 18
6 Delia Belelli R, Pistis M, Peters JA, Lambert JJ. General anaesthetic action at transmitter-gated inhibitory amino acid receptors. Tr Pharmacol Sci 1999; 20: 496502[ISI][Medline]
7 Drummond JC. Monitoring depth of anesthesia. Anesthesiology 2000; 93: 87682[ISI][Medline]
8 Angel A. Central neuronal pathways and the process of anaesthesia. Br J Anaesth 1993; 71: 14863[ISI][Medline]
9 Smythies J. The functional neuroanatomy of awareness: with a focus on the role of various anatomical systems in the control of intermodal attention. Consc Cognit 1997; 6: 45581[ISI][Medline]
10 Simos PG, Papanicolaou AC, Breier JI, et al. Insights into brain function and neural plasticity using magnetic source imaging. J Clin Neurophysiol 2000; 17: 14362[ISI][Medline]
11 Alberstone CD, Skirboll SL, Benzel EC, et al. Magnetic source imaging and brain surgery: presurgical and intraoperative planning in 26 patients. J Neurosurg 2000; 92: 7990[ISI][Medline]
12 Hodes JE, Soncrant TT, Larson DM, Carlson SG, Rapoport SI. Selective changes in local cerebral glucose utilization induced by phenobarbital in the rat. Anesthesiology 1985; 63: 6339[ISI][Medline]
13 Maekawa T, Tommasino C, Shapiro HM, Keifer-Goodman J, Kohlenberger RW. Local cerebral blood flow and glucose utilization during isoflurane anesthesia in the rat. Anesthesiology 1986; 65: 14451[ISI][Medline]
14 Ori C, Dam M, Pizzolato G, Battistin L, Giron G. Effects of isoflurane anesthesia on local cerebral glucose utilization in the rat. Anesthesiology 1986; 65: 1526[ISI][Medline]
15 Dam M, Ori C, Pizzolato G, et al. The effects of propofol anesthesia on local cerebral glucose utilization in the rat. Anesthesiology 1990; 73: 499505[ISI][Medline]
16 Lenz C, Rebel A, van Ackern K, Kuschinsky W, Waschke KF. Local cerebral blood flow, local cerebral glucose utilization, and flow-metabolism coupling during sevoflurane versus isoflurane anesthesia in rats. Anesthesiology 1998; 89: 14808[ISI][Medline]
17 Cavazzuti M, Porro CA, Biral GP, Benassi C, Barbieri GC. Ketamine effects on local cerebral blood flow and metabolism in the rat. J Cereb Blood Flow Metab 1987; 7: 80611[ISI][Medline]
18 Burdett NG, Menon DK, Carpenter TA, Jones JG, Hall LD. Visualisation of changes in regional cerebral blood flow (rCBF) produced by ketamine using long TE gradient-echo sequences: preliminary results. Magn Reson Imag 1995; 13: 54953[ISI][Medline]
19 Howseman AM, Bowtell RW. Functional magnetic resonance imaging: imaging techniques and contrast mechanisms. Phil Trans Biol Sci 1999; 354: 117994[ISI]
20 Ugurbil K, Chen W, Zhu X-H, Kim S-G, Georgopoulos A. Functional mapping in the human brain using high magnetic fields. Phil Trans: Biol Sci 1999; 354: 1195213[ISI]
21 Hansen TD, Warner DS, Todd MM, Vust LJ. The role of cerebral metabolism in determining the local cerebral blood flow effects of volatile anesthetics: evidence for persistent flow-metabolism coupling. J Cereb Blood Flow Metab 1989; 9: 3238[ISI][Medline]
22 Antognini JF, Buonocore MH, Disbrow EA, Carstens E. Isoflurane anesthesia blunts cerebral responses to noxious and innocuous stimuli: a fMRI study. Life Sci 1997; 61: PL34954[ISI]
23 Saha GB, MacIntyre WJ, Go RT. Radiopharmaceuticals for brain imaging. Semin Nucl Med 1994; 24: 32449[ISI][Medline]
24 Hoffman JM, Coleman RE. Perfusion quantitation using positron emission tomography. Invest Radiol 1992; 27: S226
25 Jones T. The role of positron emission tomography within the spectrum of medical imaging. Eur J Nucl Med 1996; 23: 20711[ISI][Medline]
26 Frackowiak RS, Friston KJ. Functional neuroanatomy of the human brain: positron emission tomography a new neuroanatomical technique. J Anat 1994; 184: 21125[ISI][Medline]
27 Jueptner M, Weiller C. Review: does measurement of regional cerebral blood flow reflect synaptic activity? Implications for PET and fMRI. Neuroimage 1995; 2: 14856[ISI][Medline]
28 Friston KJ, Price CJ, Fletcher P, Moore C, Frackowiak RS, Dolan RJ. The trouble with cognitive subtraction. Neuroimage 1996; 4: 97104[ISI][Medline]
29 Price CJ, Friston RJ. Cognitive conjunction: a new approach to brain activation experiments. Neuroimage 1997; 5: 26170[ISI][Medline]
30 Alkire MT, Pomfrett CJ, Haier RJ, et al. Functional brain imaging during anesthesia in humans: effects of halothane on global and regional cerebral glucose metabolism. Anesthesiology 1999; 90: 7019[ISI][Medline]
31 Alkire MT, Haier RJ, Shah NK, Anderson CT. Positron emission tomography study of regional cerebral metabolism in humans during isoflurane anesthesia. Anesthesiology 1997; 86: 54957[ISI][Medline]
32 Alkire MT, Haier RJ, Barker SJ, et al. Cerebral metabolism during propofol anesthesia in humans studied with positron emission tomography. Anesthesiology 1995; 82: 393403[ISI][Medline]
33 Fiset P, Paus T, Daloze T, et al. Brain mechanisms of propofol-induced loss of consciousness in humans: a positron emission tomographic study. J Neurosci 1999; 19: 550613
34 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: 110617[ISI][Medline]
35 Alkire MT, Haier RJ. Correlating in vivo anaesthetic effects with ex-vivo receptor density supports a GABAergic mechanism of action for propofol, but not for isoflurane. Br J Anaesthesia 2001; 86: 61826
36 Braestrup C, Albrechtsen R, Squires RF. High densities of benzodiazepine receptors in human cortical areas. Nature 1977; 269: 7024[ISI][Medline]
37 Zezula J, Cortes R, Probst A, Placacios JM. Benzodiazepine receptor sites in the human brain: Autoradiographic mapping. Neuroscience 1988; 25: 77195[ISI][Medline]
38 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. Consc Cognit 2000; 9: 37086
39 Matta BF, Heath KJ, Tipping K, Summors AC. Direct cerebral vasodilatory effects of sevoflurane and isoflurane. Anesthesiology 1999; 91: 67780[ISI][Medline]
40 Tsukada H, Kakiuchi T, Ando I, Shizuno H, Nakanishi S, Ouchi Y. Regulation of cerebral blood flow response to somatosensory stimulation through the cholinergic system: a positron emission tomography study in unanesthetized monkeys. Brain Res 1997; 749: 107[ISI][Medline]
41 Parola AL, Yamamura HI, Laird HE 3rd. Peripheral-type benzodiazepine receptors. Life Sci 1993; 52: 132942[ISI][Medline]
42 Frey KA, Holthoff VA, Koeppe RA, et al. Parametric in vivo imaging of benzodiazepine receptor distribution in human brain. Ann Neurol 1991; 30: 66372[ISI][Medline]
43 Ableitner A, Herz A. Influence of meprobamate and phenobarbital upon local cerebral glucose utilzation: parallelism with effects of the anxiolytic diazepam. Brain Res 1987; 403: 828
44 Kelly PA, McCulloch J. Effects of the putative GABAergic agonists, muscimol and THIP, upon local cerebral glucose utilisation. J Neurochem 1982; 39: 61324[ISI][Medline]
45 Duncan JS. Positron emission tomography receptor studies. Adv Neurol 1999; 79: 8939[Medline]
46 Jones AK, Kitchen ND, Watabe H, et al. Measurement of changes in opioid receptor binding in vivo trigeminal neuralgic pain using [11C]diprenorphine and positron emission tomography. J Cereb Blood Flow Metab 1999; 19: 8038[ISI][Medline]
47 Casey KL, Svensson P, Morrow TJ, et al. Selective opiate modulation of nociceptive processing in the human brain. J Neurophysiol 2000; 84: 52533
48 Zubieta JK, Dannals RF, Frost JJ. Gender and age influences on human brain mu-opioid receptors by PET. Am J Psychiatr 1999; 156: 8428
49 Yoshida T, Kuwabara Y, Sasaki M, et al. Sex-related differences in the muscarinic acetylcholinergic receptor human brain a positron emission tomography study. Ann Nucl Med 2000; 14: 97101[ISI][Medline]
50 Zubieta JK, Koeppe RA, Mulholland GK, Kuhl DE, Frey KA. Quantification of muscarinic cholinergic receptors with [11C]NMPB and positron emission tomography: method development and differentiation of tracer delivery from receptor binding. J Cereb Blood Flow Metab 1998; 18: 61931[ISI][Medline]
51 Grunwald F, Biersack HJ, Kuschinsky W. Nicotine receptor mapping. Eur J Nucl Med 1996; 23: 10124[ISI][Medline]