- Email: [email protected]
Effects of nitrous oxide sedation on resting electroencephalogram topography
Clinical Neurophysiology 124 (2013) 417–423
Contents lists available at SciVerse ScienceDirect
Clinical Neurophysiology journal homepage: www.elsevier.com/locate/clinph
Effects of nitrous oxide sedation on resting electroencephalogram topography Brett L. Foster a,⇑, David T.J. Liley b a b
Department of Neurology and Neurological Sciences, School of Medicine, Stanford University, Stanford, CA, United States Brain and Psychological Sciences Research Centre, Swinburne University of Technology, Melbourne, Victoria, Australia
a r t i c l e
i n f o
Article history: Accepted 14 August 2012 Available online 8 September 2012 Keywords: EEG Anesthesia Nitrous oxide Frontal theta Topography NMDA receptor
h i g h l i g h t s Unlike common inductive general anesthetics, nitrous oxide reduces the amplitude of the EEG (total power). Topographically, nitrous oxide induced EEG power suppression chiefly reflects reduced frontal delta power. During nitrous oxide washout, the EEG displays a withdrawal response of enhanced frontal theta power.
a b s t r a c t Objective: To quantify the effects of nitrous oxide (N2O) gas on electroencephalogram (EEG) topography in healthy male participants. Methods: Healthy male participants were administered 20% (n = 8) or 40% (n = 8) N2O while having highdensity (modified 10–20) noise minimized EEG recordings. Results: Nitrous oxide was found to produce clear reductions in resting total power, particularly at frontal-vertex sites. These reductions were found to principally reflect reductions in band-limited delta power. Following the termination of N2O inhalation, during N2O washout, selective increases in frontal theta power were observed that increased above baseline values. Conclusions: Nitrous oxide does not produce the classical anteriorization of slow wave activity typically seen during anesthetic induction. Instead N2O reduces frontal slow wave (delta) activity, which during gas washout produces a withdrawal response of enhanced frontal slow wave (theta) activity. Significance: Attempts to characterize a unitary mechanism of loss of consciousness during anesthesia on the basis of the topographic electroencephalographic changes is challenged by the distinct EEG effects that N2O has when compared to other well known anesthetic agents that include propofol and sevoflurane. Ó 2012 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
1. Introduction During general anesthetic induction and emergence the human electroencephalogram (EEG) undergoes a range of dose and agent dependent changes (Bennett et al., 2009; Hardmeier et al., 2012). Despite great pharmacological diversity, most inductive general anesthetics (GAs), which include the halogenated volatile agents such as sevoflurane and intravenous agents such as propofol, suppress neural activity principally through the enhancement of gamma-amino-butyric acid (GABA) action at GABAA postsynaptic receptors (Rudolph and Antkowiak, 2004). This pharmacology is also shared by less potent sedative agents such as the benzodiaze⇑ Corresponding author. Address: Department of Neurology and Neurological Sciences, School of Medicine, Stanford University, 300 Pasteur Drive, Mail Room A343, Stanford, CA 94305-5235, United States. Tel.: +1 650 757 4294. E-mail address: [email protected] (B.L. Foster).
pines and alcohol. By variably potentiating GABAA based inhibition (e.g. enhancement of the inhibitory post synaptic potential; IPSP), these GAs ameliorate the low amplitude desynchronized EEG activity associated with the awake state and induce larger amplitude slow wave oscillations more generally reflective of the states observed during deep sleep and coma (Franks, 2008), but with several important differences (Brown et al., 2010). Topographically, inductive GAs generally promote increased low-frequency delta (0–4 Hz) activity in frontal regions. Such a ‘‘frontal predominance’’ of activity, first reported in monkeys (Tinker et al., 1977) is now a commonly reported finding in the electroencephalographic characterization of anesthetic action in humans. For example John et al. (2001) found that anteriorization of frontal absolute EEG power was consistently found at loss of consciousness in patients anesthetized with a range of inductive GAs that included propofol and the inhalational agents isoflurane, desflurane and sevoflurane. Furthermore, high-density EEG studies involving
1388-2457/$36.00 Ó 2012 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.clinph.2012.08.007
418
B.L. Foster, D.T.J. Liley / Clinical Neurophysiology 124 (2013) 417–423
source modeling (Murphy et al., 2011) have shown that sleep and loss of consciousness induced by propofol are both associated with increased frontal slow wave activity, suggesting that alterations in the topography of slow wave EEG architecture is possibly key to understanding the genesis of alterations in conscious state. However, complicating such an interpretation is the fact that the electroencephalographic behavior of dissociative anesthetic agents such as nitrous oxide (N2O) and ketamine is typically not associated with the induction of slow wave activity. For an anesthetic/sedative agent N2O has peculiar effects on the EEG. Unlike inductive GAs N2O is widely believed to enhance high frequency EEG activity, which is typically associated with the alert and aroused state. For example, Yamamura et al. (1981) reported that N2O produced fast oscillatory activity, defined by a 34 Hz spectral peak, while reducing alpha power. Enhancement of higher frequency spectral power was also reported by Rampil et al. (1998), who described elevated power in hi-beta (40–50 Hz) and low-EMG (70–110 Hz) spectral bands. Rampil et al. (1998) reported paradoxical changes in the commonly used bispectral index (BIS), whereby increasing concentrations of N2O were associated with higher BIS scores and therefore higher estimated levels of awareness, whereas decreasing concentrations (e.g. during withdrawal) of N2O resulted in a reduction of BIS score and therefore lower estimated levels of awareness. However in contrast to these results Foster and Liley (2011) found that N2O inhalation in healthy males was associated with a suppression of power in the delta (0– 4 Hz) and theta (4–8 Hz) bands, while the distribution of power in higher frequency bands was largely unaffected. In either case such effects are argued to account for the inability of many processed EEG depth of anesthesia monitors to meaningfully detect the behavioral effects of N2O, as such monitors typically rely on characterizations that involve either quantifying reductions in high frequency, and/or increases in low frequency, EEG activity that attend anesthetic action. One additional difference of note is that GAs produce transient enhancement of higher frequency power during the induction, and also emergence, of anesthesia. The EEG captures this ‘paradoxical arousal’ or ‘disinhibition’ as a biphasic increase in band limited power, particularly over the alpha/beta band range (Kuizenga et al., 1998, 2001). This phenomenon occurs rapidly at the onset of pharmacological effect and will typically be replaced by the onset of slow wave and burst activity as the full potency of the anesthetic is achieved (Brown et al., 2011). In light of this, it is interesting to note that the less potent benzodiazepines, such as alprazolam and diazepam, produce a similar enhancement of the beta rhythm during sedative and anxiolytic doses, but do not produce further shifts to slow wave oscillations (Jensen et al., 2005). By comparison N2O does not appear to produce a biphasic response with the onset of inhalation, but does however show a clear rebound effect during cessation of inhalation (Foster and Liley, 2011). While the behavior of N2O on slow wave electroencephalographic activity does not accord with that of the more clinically ubiquitous inductive GAs, it nevertheless remains an open question whether it results in a similar redistribution of topographic activity to that of frontal dominance. Additionally, is there any topographical redistribution of power within specific frequency bands during N2O inhalation? Here we show, in clear distinction to that reported with inductive GAs and benzodiazepines, that sedative levels of N2O result in a reduction in frontal-midline EEG power such that it becomes occipitally dominant, and thus ‘‘posteriorized’’. The implications of these results are that a unitary mechanism for the hypnotic action of anesthetic agents based on canonical electroencephalographic changes is not a desideratum yet within sight. Instead the patterns of induced electroencephalographic activity are hypothesized to be dependent on the neuropharmacological profile of the individual anesthetic agent and
their subsequent effects electrorhythmogenesis.
upon
the
mechanisms
of
brain
2. Methods 2.1. Subjects Twenty healthy male participants (m age = 26 yrs; weight = 72 kg; height = 176 cm) were recruited for the study described below, which was approved by the Swinburne University of Technology Human Research Ethics Committee. Female participants were excluded due to an increased likelihood of nausea and emesis with N2O and difficulties in controlling for the effects of menstrual cycle and contraceptive pill on resting EEG (Kaneda et al., 1997; SolisOrtiz et al., 1994). All subjects were required to pass a general medical examination, performed by a registered physician, and to provide written voluntary informed consent for their participation. 2.2. Procedure The experimental protocol involved the inhalation of a N2O/O2 gas mixture during multichannel EEG recordings in a noiseshielded laboratory as previously reported (Foster and Liley, 2011). Using a between groups study design, participants were only subject to a single N2O condition to minimize attrition due to side effects that principally included nausea and vomiting. Prior to the recording session each participant was randomly allocated to one of the three N2O conditions: 20%, 40% or 60% inspired N2O respectively. Due to participant attrition the 60% condition was discontinued, with only data from the 20% and 40% conditions being reported here. Testing sessions commenced at 9 am and ran for approximately 2-h. Participants were asked to fast for 8-h prior to participation. Before gas inhalation recordings a 5-min resting baseline was recorded, during which the subject performed a low difficulty auditory continual performance task (aCPT, see Section 2.2.3) with the eyes closed. Recordings of EEG during N2O inhalation were 20min in duration. This period was comprised of a 5-min equilibration phase, during which the desired concentration of gas mixture was administered. This fixed duration allowed end-tidal N2O concentrations to equilibrate. After this 5-min period a marker was placed in the EEG trace to signify the beginning of a 10-min period of continuous equilibrated gas flow. After this 10-min period had elapsed (15-min in total since the commencement of N2O inhalation) N2O was discontinued with pure O2 (100%) administered for an additional 5-min, as is common for conscious sedation using N2O (Clark and Brunick, 2003). Participants performed the same aCPT during this entire 20-min recording. 2.2.1. Breathing circuit Medical grade N2O and O2 were administered through a standard closed system non-rebreathing circuit. The depressurized and flow regulated gas mixture was delivered to the participant via a 1.5 m Bain co-axial circuit. The expired gas was first collected in a 2 L reservoir, then pressure forced through a one-way valve into a large capacity Douglas bag attached to the rear of the gas cylinder cart. Collected N2O was disposed via atmospheric release post recording. Disposable anesthetic filters were attached to the participant end of the breathing circuit, making all items before the filter reusable. New disposable masks and filters were used for each participant. 2.2.2. Physiological monitoring Oxygen saturation, as well as heart rate, was obtained using a standard clinical pulse oximeter finger clip. Expired gas was ana-
B.L. Foster, D.T.J. Liley / Clinical Neurophysiology 124 (2013) 417–423
lyzed online using a Normocap 200™ (Datex-Ohmeda, GE Healthcare, USA) infra-red gas analyzer in order to determine the end tidal concentrations of N2O, O2 and CO2. These concentrations were acquired in order to monitor ongoing respiratory function and to also ensure N2O equilibration. Continuous concentration estimates were achieved through extracting a minor volume of expired gas from the breathing circuit via a small sampling line attached to the anesthetic filter, providing continual and real time concentration analysis. Data values for both the pulse oximeter and gas analyzer were manually logged once every minute during the gas condition by the experimenter. 2.2.3. Auditory continual performance task (aCPT) In order to monitor behavioral state and limit the inter-participant variability of spontaneous resting EEG, participants were asked to perform an auditory continuous performance task (aCPT). This task aimed at maintaining a basic level of resting awareness, and therefore a more comparable baseline across participants (Knott, 2000). Additionally, the expected psychoactive effects of N2O, in particular laughter and gregariousness, were hoped to be less intrusive to study protocol and EEG quality given this procedural demand on participants. Participants were presented with two auditory tones of differing frequency (1 and 2 kHz respectively) and fixed stereo amplitude (70 dB). Participants were asked to provide a left or right button response for the low (1 kHz) and high (2 kHz) pitch tones respectively. Stimulus presentation was performed using Stim 4.3™ (Neuroscan, Compumedics Ltd., Melbourne, Australia). Auditory tones were presented randomly, with equal probability and interstimulus interval of 2.5 s. Participants continuously performed the task during a baseline recording (5 min, 125 trials) and throughout the gas inhalation (20 min, 500 trials). Because task performance was expected to degrade with increasing levels of sedation, the aCPT task additionally acted as a behavioral measure of sedation (by combining accuracy and response time). Targets and responses (correctness & latency) were automatically logged for each trial. 2.2.4. EEG recording EEG acquisition was performed using two coupled 32-channel SynAmps™ EEG systems (NeuroScan). Sixty-two channels were dedicated to the acquisition of scalp EEG (Quick-Cap™, NeuroScan), whilst the remaining 2 channels were used for auxiliary electrooculogram (EOG) and electromyogram (EMG) recording. The 62-channel EEG montage was positioned according to the extended international 10:20 system, with a linked mastoids reference. EOG activity was recorded from a monopolar electrode positioned proximal to the left lateral canthus. EMG was recorded from a submental bipolar electrode pair, placed inferior to the chin on the left hand side following the sternocleidomastoid. Electrode impedance was maintained below 10 kX for all recordings. EEG was acquired continuously using Scan 4.3™ (NeuroScan) with a sample rate of 500 Hz, bounded by a digital band pass filter (0.1– 70 Hz), in a noise minimized recording laboratory. Any extraneous or biological signal artifact was dealt with offline. 2.3. EEG processing EEG data was analyzed offline using custom routines in MATLAB™ (The Mathworks, Natick, USA). Raw 62-channel EEG was initially band-pass filtered using a two-way least squares finite impulse response (FIR) band pass filter (1–40 Hz). Based on previous reports of scalp EMG power during anesthesia (Sleigh, et al. 2001), N2O’s potentiation of EMG power (Rampil et al., 1998), along with consideration of recent findings which suggest EEG power above 30 Hz can be greatly accounted for by EMG (Whitham
419
et al., 2008, 2007), the low pass filter was set to 40 Hz. Filtered EEG data was then inspected by eye for artifact, with any erroneous or failed channels being removed. Because of the 40 Hz low pass filter no specific attempt was made to remove 50 Hz mains line interference. Any remaining artifact such as eye blinks were successfully removed using independent components analysis (ICA) as implemented in the EEGLAB toolbox (Delorme and Makeig, 2004; Fitzgibbon et al., 2007; Jung et al., 2000). For quantitative analysis spectral power was calculated for the broad total power range (1–40 Hz) as well as the classical EEG bands delta (1–3 Hz), theta (4–7 Hz) alpha (8–14 Hz), beta (15– 30 Hz) and gamma (30–40 Hz). Spectral power was estimated using a fast Fourier transform of 2 s 50% overlapping Hamming windows and integrating the power across the frequency range of interest. To study the effects of N2O on EEG topography the data recording was broken into three periods of interest. Rest (baseline), was defined by the pre-gas 5-min resting period. For comparison with N2O effects rest data was compared against the final 5-min (i.e. 600–900 s) of the respective gas condition (reflecting peak effect). For comparison with N2O washout, data was taken from the N2O cessation period reflecting the last 3-min of recording, after removing 30 s from data end to avoid transients arising from terminating the EEG recording.
3. Results Across subjects N2O sedation produced reductions in the mean accuracy (correct: 97.8%; 87.8%) and extended response times (latency increase: +130 ms; +380 ms) from rest in a dose dependent manner (20%; 40% N2O respectively). Typically, one of the most commonly reported features of anesthetic action on the EEG is the enhancement of amplitude, which typically reflects the emergence of slow wave activity. Therefore we quantified the topographic changes in total EEG power (1–40 Hz) from rest to peak gas as defined above. Fig. 1 shows the topographic distribution of total power at rest and peak gas for 20% and 40% N2O inhalation. Differences between these two states are also displayed as a percentage change in power and its associated t-statistic (Fig. 1). As is typically observed total power is topographically maximal along the longitudinal midline, particularly at the occiput and vertex. From this resting distribution, N2O produces a reduction in total power, particularly at frontal-vertex sites, for both 20% and 40% conditions. Quantitatively the reduction is greater for the 40% N2O inhalation (Fig. 1). Reductions in total power reflect a decrease in the integrated spectral power. To identify unique spectral changes, we then quantified topographic changes in band limited power. As shown in Fig. 2 inhalation of 20% N2O was associated with only a reduction in delta band power, which was maximally reduced at frontal-vertex sites. As also shown in Fig. 2, inhalation of 40% N2O was associated with a reduction in delta band power, but to a greater extent at frontal-vertex sites. Thus delta band power became relatively more dominant occipitally. Changes in other bands were relatively minor and topographically more diffuse. Together these data suggest the chief component of total power reduction produced by N2O is the suppression of delta band activity, contrary to that typically observed with GAs. In light of these clear dose-dependent effects of N2O on EEG topography, we then compared the topographic changes in band limited power for the transition from peak gas to gas washout. Interestingly, for both 20% and 40% N2O inhalation, frontal-vertex theta band power was greatly enhanced during washout (Fig. 3). Changes in other bands were relatively minor for 20% N2O, but de-
420
B.L. Foster, D.T.J. Liley / Clinical Neurophysiology 124 (2013) 417–423
Rest
Peak Gas
Diff. Gas - Rest (%)
Diff. Gas - Rest (t)
150 100 50
250 200 150 100 50
20 0 −20
ΔPower (t)
200
ΔPower (%)
250
Total Power (µV 2)
Total Power (µV 2)
20% [N2O] n=8 3 2 1 0 −1 −2 −3
40% [N2O] n=8
Fig. 1. Topographic changes in EEG total power from rest to peak gas. Topographic maps for 20% (top row) and 40% (bottom row) N2O conditions. Maps show total power (lV2) at rest, peak gas and the difference between states (gas-rest) as percentage change and t-statistic (columns left to right respectively).
Fig. 2. Topographic changes in EEG band limited power from rest to peak gas. Topographic maps for 20% and 40% N2O condition. Maps show the difference between states (gas-rest) for each frequency band of interest (columns left–right) as percentage change (top row) and t-statistic (bottom row) for each gas concentration separately.
creases were observed for delta, beta and gamma bands for the 40% N2O condition. In light of these findings we then directly assessed the degree of anterior–posterior difference in power change by comparing (Wilcoxon rank sum) the significance of change (p-value maps) between anterior (all 10–20 sites anterior to, but not including, Cz) and posterior (all 10–20 sites posterior to, but not including, Cz) electrodes. Simply stated, we tested the null hypothesis that the significance of power change was not topographically different between anterior and posterior electrodes. Consistent with the topographic trends in Figs. 2 and 3, for the 20% N2O condition we observed only a significant anterior–posterior difference in power change for the delta band (p < 0.01; all other bands p > 0.05) during inhalation. Conversely, as shown in Fig. 3, during washout for the 20% condition we observed only a significant anterior–posterior difference in power change for the theta band (p < 0.05; all other bands p > 0.05). For the 40% N2O condition we also observed only a significant anterior–posterior differ-
ence in power change for the delta band (p < 0.0001; all other bands p > 0.05) during inhalation. Similarly, during washout for the 40% condition we also observed a significant anterior–posterior difference in power change for the theta band (p < 0.001; all other bands p > 0.05).
4. Discussion In this study we showed that, contrary to expectation with inductive anesthetic agents such as propofol and sevoflurane (Breshears et al., 2010; Cimenser et al., 2011; Gugino et al., 2001; Murphy et al., 2011) or sedative benzodiazepines such as diazepam (Jensen et al., 2005) that N2O action at sedative inhaled concentrations was not associated with an increase in frontal slow wave activity. In fact quite the opposite trend was found – N2O action was associated with a reduction in absolute frontal electroencephalographic power driven predominantly by reductions in delta
B.L. Foster, D.T.J. Liley / Clinical Neurophysiology 124 (2013) 417–423
421
Fig. 3. Topographic changes in EEG band limited power from peak gas to washout. Topographic maps for 20% and 40% N2O condition. Maps show the difference between states (washout-gas) for each frequency band of interest (columns left–right) as percentage change (top row) and t-statistic (bottom row) for each gas concentration separately.
band activity (Fig. 2) that rebounded, particularly in the theta band range, above baseline values following gas washout (Fig. 3). In contrast, the topographic distribution of power in the other bands during N2O action showed no consistent changes with respect to baseline distributions. Indeed the relative distribution of power in theta (4–8 Hz), alpha (8–13 Hz), beta (13–30 Hz) and gamma (>40 Hz) bands was maintained subsequent to N2O inhalation which contrasts with the effects of well-known inductive GAs such as propofol and sevoflurane, and in particular the commonly observed enhancement of beta activity during benzodiazepine action. For example, the initial onset of loss of consciousness induced using propofol is associated with a diffuse increase in low beta and high alpha band activity that eventually colligates into sustained high amplitude frontal alpha activity at higher drug concentrations (Ching et al., 2010; Feshchenko et al., 2004; Murphy et al., 2011). In a similar, yet scaled fashion, benzodiazepines enhance beta power at frontal and vertex sites, particularly for intravenous administrations (Hardmeier et al., 2012). As previously discussed (Foster and Liley, 2011), these differences in frontal EEG effects across anesthetics agents are of particular note given the common usage of frontal recording montages in depth of anesthesia monitoring. N2O washout was however associated with changes in power bands other than delta. The termination of N2O inhalation was associated with a rebound above baseline values, in absolute theta power and total power that is largely restricted to frontal locations. Such results are in-line with previous reports of N2O ‘‘withdrawal excitability’’ (Henrie et al., 1961; Rampil et al., 1998; Williams et al., 1984). For example, Williams et al. (1984) showed that N2O inhalation produced reductions in total EEG power that were subsequently elevated above baseline values following the administration of 100% O2, whereas Rampil et al. (1998), in a subset of participants, found N2O withdrawal resulted in increases in delta power above baseline that were associated with reduced Bispectral Index values. The observation of theta fronto-vertex rebound raises an interesting connection to the well documented midline theta
rhythm which is modulated by a host of cognitive paradigms particularly relevant to memory function (Mitchell et al., 2008). As the sedative levels of inhalation employed here also induce minor amnestic states, the suppression of theta power may be indicative of this influence on cognition, with the observed rebound response of frontal-midline theta reflecting the restoration of this oscillatory mode. Given that N2O action is associated with reductions in frontal delta band power we might reasonably conclude some form of EEG activation has taken place and thus posit that N2O effects will not be associated with decreases in cerebral metabolic rate. Indeed studies in animals (Pelligrino et al., 1984) and humans (Deutsch and Samra, 1990; Field et al., 1993; Reinstrup and Messeter, 1994; Reinstrup et al., 1994, 2008) show that cerebral metabolic rate (labeled glucose metabolism) is either increased or remains unchanged. In contrast, general anesthetic agents and benzodiazepines are typically reported to reduce cerebral metabolism (Alkire et al., 1995; Kaisti et al., 2003; Veselis et al., 1997), suggesting that their effects in disrupting the global brain networks associated with consciousness are quite different to those of N2O. Recent network level analyses of anesthetic action have revealed that the functional topology of cortical neuronal networks to be significantly reconfigured subsequent to the hypnotic actions of propofol or sevoflurane (Boveroux et al., 2010; Breshears et al., 2010; Cimenser et al., 2011; Deshpande et al., 2010; Ku et al., 2011; Lee et al., 2011, 2010). For example, Cimenser et al. (2011) found, using 64 channel EEG recorded during target controlled propofol infusions, that frontal increases in delta and alpha band activity were associated with strongly coordinated alpha band activity that emerged occipitally to become frontally dominant at loss of consciousness. Lee et al. (2010, 2011) found that propofol not only altered the global topology of functional delta band networks but also selectively inhibited fronto-parietal feedback activity in response to perturbations of local neuronal circuits within posterior parietal cortex. Such alterations in fronto-parietal connectivity are
422
B.L. Foster, D.T.J. Liley / Clinical Neurophysiology 124 (2013) 417–423
commonly reported during network level analyses during propofol anesthesia (Ku et al., 2011; Boveroux et al., 2010). Therefore, given N2O’s differing topographic effects on frontal delta band power it can be reasonably speculated that its action will be associated with different network level effects compared to propofol and sevoflurane. If this were subsequently shown to be the case then this would represent a significant challenge to the emerging view that there are unitary neurobiological mechanisms of anesthetic induced sedation and unconsciousness. These observed electroencephalographic differences between N2O and propofol/sevoflurane action might of course be attributable to their differing molecular/cellular sites of action and their pharmacological potency. N2O is reported to predominantly antagonize NMDA receptor activity, whereas propofol and sevoflurane potently potentiate GABAergic receptor activity, as do the benzodiazepines (Rudolph and Antkowiak, 2004). However such a simple explanation is complicated by the fact that xenon, which is also reported to significantly antagonize NMDA receptor activity and to have no GABAergic action, has quite different electroencephalographic properties. Xenon has been demonstrated to slow the EEG and produce an increase in total power especially in frontal regions (Johnson et al., 2003). In particular, xenon has been associated with an increase in slow wave theta activity (Hartmann et al., 1991; Laitio et al., 2008; Utsumi et al., 1998). This suggests contrasting mechanisms of action compared with N2O. An additional consideration is the clear differences in the potency of these agents, with N2O not capable of producing full anesthesia even at high concentrations (>60%), whereas Xenon and the other inductive GABAergic anesthetics show a far greater potency sufficient for surgical anesthesia. Therefore, future research needs to better characterize the dynamical and network level differences between N2O, xenon and propofol/sevoflurane as this will help clarify the plausibility of pursuing unitary mechanistic descriptions of anesthetic action on the basis of macroscopic brain activity. Furthermore, such studies will need to address the inherent limitations of our experimental methodology in which participants were exposed to only one steady-state N2O level. In order to control for inter-subject variability it will be important that subsequent studies involve, where possible, larger sample size repeated measure crossover designs in which multiple agents and concentrations are administered. 5. Financial interests This research was supported by departmental funds to (D.T.J.L.). B.L.F has previously (2009) been paid as a scientific consultant by Cortical Dynamics Ltd., an unlisted subsidiary of Biopharmica Ltd. (Perth, Australia), as part of the development of an EEG based depth of anesthesia monitor. D.T.J.L has an unvalued equity stake in Cortical Dynamics Ltd. D.T.J.L is an inventor on 4 patent applications filed by Cortical Dynamics Ltd. since 2004 that describe new approaches to monitoring depth of anesthesia using the EEG. To date only one patent in one territory (NZ) has been issued. None of the IP declared in any published, pending or granted patent has been licensed. References Alkire MT, Haier RJ, Barker SJ, Shah NK, Wu JC, Kao YJ. Cerebral metabolism during propofol anesthesia in humans studied with positron emission tomography. Anesthesiology 1995;82:393–403. Bennett C, Voss LJ, Barnard JP, Sleigh JW. Practical use of the raw electroencephalogram waveform during general anesthesia: the art and science. Anesth Analg 2009;109:539–50. Boveroux P, Vanhaudenhuyse A, Bruno MA, Noirhomme Q, Lauwick S, Luxen A, et al. Breakdown of within- and between-network resting state functional magnetic resonance imaging connectivity during propofol-induced loss of consciousness. Anesthesiology 2010;113:1038–53.
Breshears JD, Roland JL, Sharma M, Gaona CM, Freudenburg ZV, Tempelhoff R, et al. Stable and dynamic cortical electrophysiology of induction and emergence with propofol anesthesia. Proc Natl Acad Sci USA 2010;107:21170–5. Brown EN, Lydic R, Schiff ND. General anesthesia, sleep, and coma. N Engl J Med 2010;363:2638–50. Brown EN, Purdon PL, Van Dort CJ. General anesthesia and altered states of arousal: a systems neuroscience analysis. Annu Rev Neurosci 2011;34:601–28. Ching S, Cimenser A, Purdon PL, Brown EN, Kopell NJ. Thalamocortical model for a propofol-induced alpha-rhythm associated with loss of consciousness. Proc Natl Acad Sci USA 2010;107:22665–70. Cimenser A, Purdon PL, Pierce ET, Walsh JL, Salazar-Gomez AF, Harrell PG, et al. Tracking brain states under general anesthesia by using global coherence analysis. Proc Natl Acad Sci USA 2011;108:8832–7. Clark MS, Brunick AL. Handbook of nitrous oxide and oxygen sedation. 2nd ed. St Louis, Mo: Mosby; 2003. Delorme A, Makeig S. EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis. J Neurosci Methods 2004;134:9–21. Deshpande G, Kerssens C, Sebel PS, Hu X. Altered local coherence in the default mode network due to sevoflurane anesthesia. Brain Res 2010;1318: 110–21. Deutsch G, Samra SK. Effects of nitrous oxide on global and regional cortical blood flow. Stroke 1990;21:1293–8. Feshchenko VA, Veselis RA, Reinsel RA. Propofol-induced alpha rhythm. Neuropsychobiology 2004;50:257–66. Field LM, Dorrance DE, Krzeminska EK, Barsoum LZ. Effect of nitrous oxide on cerebral blood flow in normal humans. Br J Anaesth 1993;70:154–9. Fitzgibbon SP, Powers DM, Pope KJ, Clark CR. Removal of EEG noise and artifact using blind source separation. J Clin Neurophysiol 2007;24:232–43. Foster BL, Liley DT. Nitrous oxide paradoxically modulates slow electroencephalogram oscillations: implications for anesthesia monitoring. Anesth Analg 2011;113:758–65. Franks NP. General anaesthesia: from molecular targets to neuronal pathways of sleep and arousal. Nat Rev Neurosci 2008;9:370–86. Gugino LD, Chabot RJ, Prichep LS, John ER, Formanek V, Aglio LS. Quantitative EEG changes associated with loss and return of consciousness in healthy adult volunteers anaesthetized with propofol or sevoflurane. Br J Anaesth 2001;87:421–8. Hardmeier M, Zimmermann R, Ruegg S, Pfluger M, Deuster S, Suter K, et al. Intranasal midazolam: pharmacokinetics and pharmacodynamics assessed by quantitative EEG in healthy volunteers. Clin Pharmacol Ther 2012;91:856–62. Hartmann A, Dettmers C, Schuier FJ, Wassmann HD, Schumacher HW. Effect of stable xenon on regional cerebral blood flow and the electroencephalogram in normal volunteers. Stroke 1991;22:182–9. Henrie JR, Parkhouse J, Bickford RG. Alteration of human consciousness by nitrous oxide as assessed electro-encephalography and psychological tests. Anesthesiology 1961;22:247–59. Jensen O, Goel P, Kopell N, Pohja M, Hari R, Ermentrout B. On the human sensorimotor-cortex beta rhythm: sources and modeling. NeuroImage 2005;26:347–55. John ER, Prichep LS, Kox W, Valdes-Sosa P, Bosch-Bayard J, Aubert E, et al. Invariant reversible QEEG effects of anesthetics. Conscious Cogn 2001;10:165–83. Johnson BW, Sleigh JW, Kirk IJ, Williams ML. High-density EEG mapping during general anaesthesia with Xenon and propofol: A pilot study. Anaesth Intensive Care 2003;31:155–63. Jung TP, Makeig S, Humphries C, Lee TW, McKeown MJ, Iragui V, et al. Removing electroencephalographic artifacts by blind source separation. Psychophysiology 2000;37:163–78. Kaisti KK, Langsjo JW, Aalto S, Oikonen V, Sipila H, Teras M, et al. 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. Kaneda Y, Ikuta T, Nakayama H, Kagawa K, Furuta N. Visual evoked potential and electroencephalogram of healthy females during the menstrual cycle. J Med Invest 1997;44:41–6. Knott VJ. Quantitative EEG methods and measures in human psychopharmacological research. Hum Psychopharmacol 2000;15:479–98. Ku SW, Lee U, Noh GJ, Jun IG, Mashour GA. Preferential inhibition of frontal-toparietal feedback connectivity is a neurophysiologic correlate of general anesthesia in surgical patients. PLoS One 2011;6:e25155. Kuizenga K, Kalkman CJ, Hennis PJ. Quantitative electroencephalographic analysis of the biphasic concentration-effect relationship of propofol in surgical patients during extradural analgesia. Br J Anaesth 1998;80:725–32. Kuizenga K, Wierda JM, Kalkman CJ. Biphasic EEG changes in relation to loss of consciousness during induction with thiopental, propofol, etomidate, midazolam or sevoflurane. Br J Anaesth 2001;86:354–60. Laitio RM, Kaskinoro K, Särkelä MOK, Kaisti KK, Salmi E, Maksimow A, et al. Bispectral index, entropy, and quantitative electroencephalogram during single-agent xenon anesthesia. Anesthesiology 2008;108:63–70. Lee U, Muller M, Noh GJ, Choi B, Mashour GA. Dissociable network properties of anesthetic state transitions. Anesthesiology 2011;114:872–81. Lee U, Oh G, Kim S, Noh G, Choi B, Mashour GA. Brain networks maintain a scale-free organization across consciousness, anesthesia, and recovery: evidence for adaptive reconfiguration. Anesthesiology 2010;113:1081–91. Mitchell DJ, McNaughton N, Flanagan D, Kirk IJ. Frontal-midline theta from the perspective of hippocampal ‘‘theta’’. Prog Neurobiol 2008;86:156–85.
B.L. Foster, D.T.J. Liley / Clinical Neurophysiology 124 (2013) 417–423 Murphy M, Bruno MA, Riedner BA, Boveroux P, Noirhomme Q, Landsness EC, et al. Propofol anesthesia and sleep: a high-density EEG study. Sleep 2011;34: 283–91. Pelligrino DA, Miletich DJ, Hoffman WE, Albrecht RF. Nitrous oxide markedly increases cerebral cortical metabolic rate and blood flow in the goat. Anesthesiology 1984;60:405–12. Rampil IJ, Kim JS, Lenhardt R, Negishi C, Sessler DI. Bispectral EEG index during nitrous oxide administration. Anesthesiology 1998;89:671–7. Reinstrup P, Messeter K. Cerebrovascular response to nitrous oxide. Acta Anaesthesiol Scand 1994;38:761–2. Reinstrup P, Ryding E, Algotsson L, Berntman L, Uski T. Effects of nitrous oxide on human regional cerebral blood flow and isolated pial arteries. Anesthesiology 1994;81:396–402. Reinstrup P, Ryding E, Ohlsson T, Sandell A, Erlandsson K, Ljunggren K, et al. Regional cerebral metabolic rate (positron emission tomography) during inhalation of nitrous oxide 50% in humans. Br J Anaesth 2008;100:66–71. Rudolph U, Antkowiak B. Molecular and neuronal substrates for general anaesthetics. Nat Rev Neurosci 2004;5:709–20. Solis-Ortiz S, Ramos J, Arce C, Guevara MA, Corsi-Cabrera M. EEG oscillations during menstrual cycle. Int J Neurosci 1994;76:279–92. Tinker JH, Sharbrough FW, Michenfelder JD. Anterior shift of the dominant EEG rhytham during anesthesia in the Java monkey: correlation with anesthetic potency. Anesthesiology 1977;46:252–9.
423
Utsumi J, Adachi T, Kurata J, Miyazaki Y, Shibata M, Murakawa M, et al. Effect of xenon on central nervous system electrical activity during sevoflurane anaesthesia in cats: Comparison with nitrous oxide. Br J Anaesth 1998;80:628–33. Veselis RA, Reinsel RA, Beattie BJ, Mawlawi OR, Feshchenko VA, DiResta GR, et al. Midazolam changes cerebral blood flow in discrete brain regions: an H2(15)O positron emission tomography study. Anesthesiology 1997;87:1106–17. Whitham EM, Lewis T, Pope KJ, Fitzgibbon SP, Clark CR, Loveless S, et al. Thinking activates EMG in scalp electrical recordings. Clin Neurophysiol 2008;119:1166–75. Whitham EM, Pope KJ, Fitzgibbon SP, Lewis T, Clark CR, Loveless S, et al. Scalp electrical recording during paralysis: quantitative evidence that EEG frequencies above 20 Hz are contaminated by EMG. Clin Neurophysiol 2007;118:1877–88. Williams DJ, Morgan RJ, Sebel PS, Maynard DE. The effect of nitrous oxide on cerebral electrical activity. Anaesthesia 1984;39:422–5. Yamamura T, Fukuda M, Takeya H, Goto Y, Furuwaka K. Fast oscillatory EEG activity by analgesic concentrations of nitrous oxide in man. Anaesth Analg 1981;60:283–8.
Contents lists available at SciVerse ScienceDirect
Clinical Neurophysiology journal homepage: www.elsevier.com/locate/clinph
Effects of nitrous oxide sedation on resting electroencephalogram topography Brett L. Foster a,⇑, David T.J. Liley b a b
Department of Neurology and Neurological Sciences, School of Medicine, Stanford University, Stanford, CA, United States Brain and Psychological Sciences Research Centre, Swinburne University of Technology, Melbourne, Victoria, Australia
a r t i c l e
i n f o
Article history: Accepted 14 August 2012 Available online 8 September 2012 Keywords: EEG Anesthesia Nitrous oxide Frontal theta Topography NMDA receptor
h i g h l i g h t s Unlike common inductive general anesthetics, nitrous oxide reduces the amplitude of the EEG (total power). Topographically, nitrous oxide induced EEG power suppression chiefly reflects reduced frontal delta power. During nitrous oxide washout, the EEG displays a withdrawal response of enhanced frontal theta power.
a b s t r a c t Objective: To quantify the effects of nitrous oxide (N2O) gas on electroencephalogram (EEG) topography in healthy male participants. Methods: Healthy male participants were administered 20% (n = 8) or 40% (n = 8) N2O while having highdensity (modified 10–20) noise minimized EEG recordings. Results: Nitrous oxide was found to produce clear reductions in resting total power, particularly at frontal-vertex sites. These reductions were found to principally reflect reductions in band-limited delta power. Following the termination of N2O inhalation, during N2O washout, selective increases in frontal theta power were observed that increased above baseline values. Conclusions: Nitrous oxide does not produce the classical anteriorization of slow wave activity typically seen during anesthetic induction. Instead N2O reduces frontal slow wave (delta) activity, which during gas washout produces a withdrawal response of enhanced frontal slow wave (theta) activity. Significance: Attempts to characterize a unitary mechanism of loss of consciousness during anesthesia on the basis of the topographic electroencephalographic changes is challenged by the distinct EEG effects that N2O has when compared to other well known anesthetic agents that include propofol and sevoflurane. Ó 2012 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
1. Introduction During general anesthetic induction and emergence the human electroencephalogram (EEG) undergoes a range of dose and agent dependent changes (Bennett et al., 2009; Hardmeier et al., 2012). Despite great pharmacological diversity, most inductive general anesthetics (GAs), which include the halogenated volatile agents such as sevoflurane and intravenous agents such as propofol, suppress neural activity principally through the enhancement of gamma-amino-butyric acid (GABA) action at GABAA postsynaptic receptors (Rudolph and Antkowiak, 2004). This pharmacology is also shared by less potent sedative agents such as the benzodiaze⇑ Corresponding author. Address: Department of Neurology and Neurological Sciences, School of Medicine, Stanford University, 300 Pasteur Drive, Mail Room A343, Stanford, CA 94305-5235, United States. Tel.: +1 650 757 4294. E-mail address: [email protected] (B.L. Foster).
pines and alcohol. By variably potentiating GABAA based inhibition (e.g. enhancement of the inhibitory post synaptic potential; IPSP), these GAs ameliorate the low amplitude desynchronized EEG activity associated with the awake state and induce larger amplitude slow wave oscillations more generally reflective of the states observed during deep sleep and coma (Franks, 2008), but with several important differences (Brown et al., 2010). Topographically, inductive GAs generally promote increased low-frequency delta (0–4 Hz) activity in frontal regions. Such a ‘‘frontal predominance’’ of activity, first reported in monkeys (Tinker et al., 1977) is now a commonly reported finding in the electroencephalographic characterization of anesthetic action in humans. For example John et al. (2001) found that anteriorization of frontal absolute EEG power was consistently found at loss of consciousness in patients anesthetized with a range of inductive GAs that included propofol and the inhalational agents isoflurane, desflurane and sevoflurane. Furthermore, high-density EEG studies involving
1388-2457/$36.00 Ó 2012 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.clinph.2012.08.007
418
B.L. Foster, D.T.J. Liley / Clinical Neurophysiology 124 (2013) 417–423
source modeling (Murphy et al., 2011) have shown that sleep and loss of consciousness induced by propofol are both associated with increased frontal slow wave activity, suggesting that alterations in the topography of slow wave EEG architecture is possibly key to understanding the genesis of alterations in conscious state. However, complicating such an interpretation is the fact that the electroencephalographic behavior of dissociative anesthetic agents such as nitrous oxide (N2O) and ketamine is typically not associated with the induction of slow wave activity. For an anesthetic/sedative agent N2O has peculiar effects on the EEG. Unlike inductive GAs N2O is widely believed to enhance high frequency EEG activity, which is typically associated with the alert and aroused state. For example, Yamamura et al. (1981) reported that N2O produced fast oscillatory activity, defined by a 34 Hz spectral peak, while reducing alpha power. Enhancement of higher frequency spectral power was also reported by Rampil et al. (1998), who described elevated power in hi-beta (40–50 Hz) and low-EMG (70–110 Hz) spectral bands. Rampil et al. (1998) reported paradoxical changes in the commonly used bispectral index (BIS), whereby increasing concentrations of N2O were associated with higher BIS scores and therefore higher estimated levels of awareness, whereas decreasing concentrations (e.g. during withdrawal) of N2O resulted in a reduction of BIS score and therefore lower estimated levels of awareness. However in contrast to these results Foster and Liley (2011) found that N2O inhalation in healthy males was associated with a suppression of power in the delta (0– 4 Hz) and theta (4–8 Hz) bands, while the distribution of power in higher frequency bands was largely unaffected. In either case such effects are argued to account for the inability of many processed EEG depth of anesthesia monitors to meaningfully detect the behavioral effects of N2O, as such monitors typically rely on characterizations that involve either quantifying reductions in high frequency, and/or increases in low frequency, EEG activity that attend anesthetic action. One additional difference of note is that GAs produce transient enhancement of higher frequency power during the induction, and also emergence, of anesthesia. The EEG captures this ‘paradoxical arousal’ or ‘disinhibition’ as a biphasic increase in band limited power, particularly over the alpha/beta band range (Kuizenga et al., 1998, 2001). This phenomenon occurs rapidly at the onset of pharmacological effect and will typically be replaced by the onset of slow wave and burst activity as the full potency of the anesthetic is achieved (Brown et al., 2011). In light of this, it is interesting to note that the less potent benzodiazepines, such as alprazolam and diazepam, produce a similar enhancement of the beta rhythm during sedative and anxiolytic doses, but do not produce further shifts to slow wave oscillations (Jensen et al., 2005). By comparison N2O does not appear to produce a biphasic response with the onset of inhalation, but does however show a clear rebound effect during cessation of inhalation (Foster and Liley, 2011). While the behavior of N2O on slow wave electroencephalographic activity does not accord with that of the more clinically ubiquitous inductive GAs, it nevertheless remains an open question whether it results in a similar redistribution of topographic activity to that of frontal dominance. Additionally, is there any topographical redistribution of power within specific frequency bands during N2O inhalation? Here we show, in clear distinction to that reported with inductive GAs and benzodiazepines, that sedative levels of N2O result in a reduction in frontal-midline EEG power such that it becomes occipitally dominant, and thus ‘‘posteriorized’’. The implications of these results are that a unitary mechanism for the hypnotic action of anesthetic agents based on canonical electroencephalographic changes is not a desideratum yet within sight. Instead the patterns of induced electroencephalographic activity are hypothesized to be dependent on the neuropharmacological profile of the individual anesthetic agent and
their subsequent effects electrorhythmogenesis.
upon
the
mechanisms
of
brain
2. Methods 2.1. Subjects Twenty healthy male participants (m age = 26 yrs; weight = 72 kg; height = 176 cm) were recruited for the study described below, which was approved by the Swinburne University of Technology Human Research Ethics Committee. Female participants were excluded due to an increased likelihood of nausea and emesis with N2O and difficulties in controlling for the effects of menstrual cycle and contraceptive pill on resting EEG (Kaneda et al., 1997; SolisOrtiz et al., 1994). All subjects were required to pass a general medical examination, performed by a registered physician, and to provide written voluntary informed consent for their participation. 2.2. Procedure The experimental protocol involved the inhalation of a N2O/O2 gas mixture during multichannel EEG recordings in a noiseshielded laboratory as previously reported (Foster and Liley, 2011). Using a between groups study design, participants were only subject to a single N2O condition to minimize attrition due to side effects that principally included nausea and vomiting. Prior to the recording session each participant was randomly allocated to one of the three N2O conditions: 20%, 40% or 60% inspired N2O respectively. Due to participant attrition the 60% condition was discontinued, with only data from the 20% and 40% conditions being reported here. Testing sessions commenced at 9 am and ran for approximately 2-h. Participants were asked to fast for 8-h prior to participation. Before gas inhalation recordings a 5-min resting baseline was recorded, during which the subject performed a low difficulty auditory continual performance task (aCPT, see Section 2.2.3) with the eyes closed. Recordings of EEG during N2O inhalation were 20min in duration. This period was comprised of a 5-min equilibration phase, during which the desired concentration of gas mixture was administered. This fixed duration allowed end-tidal N2O concentrations to equilibrate. After this 5-min period a marker was placed in the EEG trace to signify the beginning of a 10-min period of continuous equilibrated gas flow. After this 10-min period had elapsed (15-min in total since the commencement of N2O inhalation) N2O was discontinued with pure O2 (100%) administered for an additional 5-min, as is common for conscious sedation using N2O (Clark and Brunick, 2003). Participants performed the same aCPT during this entire 20-min recording. 2.2.1. Breathing circuit Medical grade N2O and O2 were administered through a standard closed system non-rebreathing circuit. The depressurized and flow regulated gas mixture was delivered to the participant via a 1.5 m Bain co-axial circuit. The expired gas was first collected in a 2 L reservoir, then pressure forced through a one-way valve into a large capacity Douglas bag attached to the rear of the gas cylinder cart. Collected N2O was disposed via atmospheric release post recording. Disposable anesthetic filters were attached to the participant end of the breathing circuit, making all items before the filter reusable. New disposable masks and filters were used for each participant. 2.2.2. Physiological monitoring Oxygen saturation, as well as heart rate, was obtained using a standard clinical pulse oximeter finger clip. Expired gas was ana-
B.L. Foster, D.T.J. Liley / Clinical Neurophysiology 124 (2013) 417–423
lyzed online using a Normocap 200™ (Datex-Ohmeda, GE Healthcare, USA) infra-red gas analyzer in order to determine the end tidal concentrations of N2O, O2 and CO2. These concentrations were acquired in order to monitor ongoing respiratory function and to also ensure N2O equilibration. Continuous concentration estimates were achieved through extracting a minor volume of expired gas from the breathing circuit via a small sampling line attached to the anesthetic filter, providing continual and real time concentration analysis. Data values for both the pulse oximeter and gas analyzer were manually logged once every minute during the gas condition by the experimenter. 2.2.3. Auditory continual performance task (aCPT) In order to monitor behavioral state and limit the inter-participant variability of spontaneous resting EEG, participants were asked to perform an auditory continuous performance task (aCPT). This task aimed at maintaining a basic level of resting awareness, and therefore a more comparable baseline across participants (Knott, 2000). Additionally, the expected psychoactive effects of N2O, in particular laughter and gregariousness, were hoped to be less intrusive to study protocol and EEG quality given this procedural demand on participants. Participants were presented with two auditory tones of differing frequency (1 and 2 kHz respectively) and fixed stereo amplitude (70 dB). Participants were asked to provide a left or right button response for the low (1 kHz) and high (2 kHz) pitch tones respectively. Stimulus presentation was performed using Stim 4.3™ (Neuroscan, Compumedics Ltd., Melbourne, Australia). Auditory tones were presented randomly, with equal probability and interstimulus interval of 2.5 s. Participants continuously performed the task during a baseline recording (5 min, 125 trials) and throughout the gas inhalation (20 min, 500 trials). Because task performance was expected to degrade with increasing levels of sedation, the aCPT task additionally acted as a behavioral measure of sedation (by combining accuracy and response time). Targets and responses (correctness & latency) were automatically logged for each trial. 2.2.4. EEG recording EEG acquisition was performed using two coupled 32-channel SynAmps™ EEG systems (NeuroScan). Sixty-two channels were dedicated to the acquisition of scalp EEG (Quick-Cap™, NeuroScan), whilst the remaining 2 channels were used for auxiliary electrooculogram (EOG) and electromyogram (EMG) recording. The 62-channel EEG montage was positioned according to the extended international 10:20 system, with a linked mastoids reference. EOG activity was recorded from a monopolar electrode positioned proximal to the left lateral canthus. EMG was recorded from a submental bipolar electrode pair, placed inferior to the chin on the left hand side following the sternocleidomastoid. Electrode impedance was maintained below 10 kX for all recordings. EEG was acquired continuously using Scan 4.3™ (NeuroScan) with a sample rate of 500 Hz, bounded by a digital band pass filter (0.1– 70 Hz), in a noise minimized recording laboratory. Any extraneous or biological signal artifact was dealt with offline. 2.3. EEG processing EEG data was analyzed offline using custom routines in MATLAB™ (The Mathworks, Natick, USA). Raw 62-channel EEG was initially band-pass filtered using a two-way least squares finite impulse response (FIR) band pass filter (1–40 Hz). Based on previous reports of scalp EMG power during anesthesia (Sleigh, et al. 2001), N2O’s potentiation of EMG power (Rampil et al., 1998), along with consideration of recent findings which suggest EEG power above 30 Hz can be greatly accounted for by EMG (Whitham
419
et al., 2008, 2007), the low pass filter was set to 40 Hz. Filtered EEG data was then inspected by eye for artifact, with any erroneous or failed channels being removed. Because of the 40 Hz low pass filter no specific attempt was made to remove 50 Hz mains line interference. Any remaining artifact such as eye blinks were successfully removed using independent components analysis (ICA) as implemented in the EEGLAB toolbox (Delorme and Makeig, 2004; Fitzgibbon et al., 2007; Jung et al., 2000). For quantitative analysis spectral power was calculated for the broad total power range (1–40 Hz) as well as the classical EEG bands delta (1–3 Hz), theta (4–7 Hz) alpha (8–14 Hz), beta (15– 30 Hz) and gamma (30–40 Hz). Spectral power was estimated using a fast Fourier transform of 2 s 50% overlapping Hamming windows and integrating the power across the frequency range of interest. To study the effects of N2O on EEG topography the data recording was broken into three periods of interest. Rest (baseline), was defined by the pre-gas 5-min resting period. For comparison with N2O effects rest data was compared against the final 5-min (i.e. 600–900 s) of the respective gas condition (reflecting peak effect). For comparison with N2O washout, data was taken from the N2O cessation period reflecting the last 3-min of recording, after removing 30 s from data end to avoid transients arising from terminating the EEG recording.
3. Results Across subjects N2O sedation produced reductions in the mean accuracy (correct: 97.8%; 87.8%) and extended response times (latency increase: +130 ms; +380 ms) from rest in a dose dependent manner (20%; 40% N2O respectively). Typically, one of the most commonly reported features of anesthetic action on the EEG is the enhancement of amplitude, which typically reflects the emergence of slow wave activity. Therefore we quantified the topographic changes in total EEG power (1–40 Hz) from rest to peak gas as defined above. Fig. 1 shows the topographic distribution of total power at rest and peak gas for 20% and 40% N2O inhalation. Differences between these two states are also displayed as a percentage change in power and its associated t-statistic (Fig. 1). As is typically observed total power is topographically maximal along the longitudinal midline, particularly at the occiput and vertex. From this resting distribution, N2O produces a reduction in total power, particularly at frontal-vertex sites, for both 20% and 40% conditions. Quantitatively the reduction is greater for the 40% N2O inhalation (Fig. 1). Reductions in total power reflect a decrease in the integrated spectral power. To identify unique spectral changes, we then quantified topographic changes in band limited power. As shown in Fig. 2 inhalation of 20% N2O was associated with only a reduction in delta band power, which was maximally reduced at frontal-vertex sites. As also shown in Fig. 2, inhalation of 40% N2O was associated with a reduction in delta band power, but to a greater extent at frontal-vertex sites. Thus delta band power became relatively more dominant occipitally. Changes in other bands were relatively minor and topographically more diffuse. Together these data suggest the chief component of total power reduction produced by N2O is the suppression of delta band activity, contrary to that typically observed with GAs. In light of these clear dose-dependent effects of N2O on EEG topography, we then compared the topographic changes in band limited power for the transition from peak gas to gas washout. Interestingly, for both 20% and 40% N2O inhalation, frontal-vertex theta band power was greatly enhanced during washout (Fig. 3). Changes in other bands were relatively minor for 20% N2O, but de-
420
B.L. Foster, D.T.J. Liley / Clinical Neurophysiology 124 (2013) 417–423
Rest
Peak Gas
Diff. Gas - Rest (%)
Diff. Gas - Rest (t)
150 100 50
250 200 150 100 50
20 0 −20
ΔPower (t)
200
ΔPower (%)
250
Total Power (µV 2)
Total Power (µV 2)
20% [N2O] n=8 3 2 1 0 −1 −2 −3
40% [N2O] n=8
Fig. 1. Topographic changes in EEG total power from rest to peak gas. Topographic maps for 20% (top row) and 40% (bottom row) N2O conditions. Maps show total power (lV2) at rest, peak gas and the difference between states (gas-rest) as percentage change and t-statistic (columns left to right respectively).
Fig. 2. Topographic changes in EEG band limited power from rest to peak gas. Topographic maps for 20% and 40% N2O condition. Maps show the difference between states (gas-rest) for each frequency band of interest (columns left–right) as percentage change (top row) and t-statistic (bottom row) for each gas concentration separately.
creases were observed for delta, beta and gamma bands for the 40% N2O condition. In light of these findings we then directly assessed the degree of anterior–posterior difference in power change by comparing (Wilcoxon rank sum) the significance of change (p-value maps) between anterior (all 10–20 sites anterior to, but not including, Cz) and posterior (all 10–20 sites posterior to, but not including, Cz) electrodes. Simply stated, we tested the null hypothesis that the significance of power change was not topographically different between anterior and posterior electrodes. Consistent with the topographic trends in Figs. 2 and 3, for the 20% N2O condition we observed only a significant anterior–posterior difference in power change for the delta band (p < 0.01; all other bands p > 0.05) during inhalation. Conversely, as shown in Fig. 3, during washout for the 20% condition we observed only a significant anterior–posterior difference in power change for the theta band (p < 0.05; all other bands p > 0.05). For the 40% N2O condition we also observed only a significant anterior–posterior differ-
ence in power change for the delta band (p < 0.0001; all other bands p > 0.05) during inhalation. Similarly, during washout for the 40% condition we also observed a significant anterior–posterior difference in power change for the theta band (p < 0.001; all other bands p > 0.05).
4. Discussion In this study we showed that, contrary to expectation with inductive anesthetic agents such as propofol and sevoflurane (Breshears et al., 2010; Cimenser et al., 2011; Gugino et al., 2001; Murphy et al., 2011) or sedative benzodiazepines such as diazepam (Jensen et al., 2005) that N2O action at sedative inhaled concentrations was not associated with an increase in frontal slow wave activity. In fact quite the opposite trend was found – N2O action was associated with a reduction in absolute frontal electroencephalographic power driven predominantly by reductions in delta
B.L. Foster, D.T.J. Liley / Clinical Neurophysiology 124 (2013) 417–423
421
Fig. 3. Topographic changes in EEG band limited power from peak gas to washout. Topographic maps for 20% and 40% N2O condition. Maps show the difference between states (washout-gas) for each frequency band of interest (columns left–right) as percentage change (top row) and t-statistic (bottom row) for each gas concentration separately.
band activity (Fig. 2) that rebounded, particularly in the theta band range, above baseline values following gas washout (Fig. 3). In contrast, the topographic distribution of power in the other bands during N2O action showed no consistent changes with respect to baseline distributions. Indeed the relative distribution of power in theta (4–8 Hz), alpha (8–13 Hz), beta (13–30 Hz) and gamma (>40 Hz) bands was maintained subsequent to N2O inhalation which contrasts with the effects of well-known inductive GAs such as propofol and sevoflurane, and in particular the commonly observed enhancement of beta activity during benzodiazepine action. For example, the initial onset of loss of consciousness induced using propofol is associated with a diffuse increase in low beta and high alpha band activity that eventually colligates into sustained high amplitude frontal alpha activity at higher drug concentrations (Ching et al., 2010; Feshchenko et al., 2004; Murphy et al., 2011). In a similar, yet scaled fashion, benzodiazepines enhance beta power at frontal and vertex sites, particularly for intravenous administrations (Hardmeier et al., 2012). As previously discussed (Foster and Liley, 2011), these differences in frontal EEG effects across anesthetics agents are of particular note given the common usage of frontal recording montages in depth of anesthesia monitoring. N2O washout was however associated with changes in power bands other than delta. The termination of N2O inhalation was associated with a rebound above baseline values, in absolute theta power and total power that is largely restricted to frontal locations. Such results are in-line with previous reports of N2O ‘‘withdrawal excitability’’ (Henrie et al., 1961; Rampil et al., 1998; Williams et al., 1984). For example, Williams et al. (1984) showed that N2O inhalation produced reductions in total EEG power that were subsequently elevated above baseline values following the administration of 100% O2, whereas Rampil et al. (1998), in a subset of participants, found N2O withdrawal resulted in increases in delta power above baseline that were associated with reduced Bispectral Index values. The observation of theta fronto-vertex rebound raises an interesting connection to the well documented midline theta
rhythm which is modulated by a host of cognitive paradigms particularly relevant to memory function (Mitchell et al., 2008). As the sedative levels of inhalation employed here also induce minor amnestic states, the suppression of theta power may be indicative of this influence on cognition, with the observed rebound response of frontal-midline theta reflecting the restoration of this oscillatory mode. Given that N2O action is associated with reductions in frontal delta band power we might reasonably conclude some form of EEG activation has taken place and thus posit that N2O effects will not be associated with decreases in cerebral metabolic rate. Indeed studies in animals (Pelligrino et al., 1984) and humans (Deutsch and Samra, 1990; Field et al., 1993; Reinstrup and Messeter, 1994; Reinstrup et al., 1994, 2008) show that cerebral metabolic rate (labeled glucose metabolism) is either increased or remains unchanged. In contrast, general anesthetic agents and benzodiazepines are typically reported to reduce cerebral metabolism (Alkire et al., 1995; Kaisti et al., 2003; Veselis et al., 1997), suggesting that their effects in disrupting the global brain networks associated with consciousness are quite different to those of N2O. Recent network level analyses of anesthetic action have revealed that the functional topology of cortical neuronal networks to be significantly reconfigured subsequent to the hypnotic actions of propofol or sevoflurane (Boveroux et al., 2010; Breshears et al., 2010; Cimenser et al., 2011; Deshpande et al., 2010; Ku et al., 2011; Lee et al., 2011, 2010). For example, Cimenser et al. (2011) found, using 64 channel EEG recorded during target controlled propofol infusions, that frontal increases in delta and alpha band activity were associated with strongly coordinated alpha band activity that emerged occipitally to become frontally dominant at loss of consciousness. Lee et al. (2010, 2011) found that propofol not only altered the global topology of functional delta band networks but also selectively inhibited fronto-parietal feedback activity in response to perturbations of local neuronal circuits within posterior parietal cortex. Such alterations in fronto-parietal connectivity are
422
B.L. Foster, D.T.J. Liley / Clinical Neurophysiology 124 (2013) 417–423
commonly reported during network level analyses during propofol anesthesia (Ku et al., 2011; Boveroux et al., 2010). Therefore, given N2O’s differing topographic effects on frontal delta band power it can be reasonably speculated that its action will be associated with different network level effects compared to propofol and sevoflurane. If this were subsequently shown to be the case then this would represent a significant challenge to the emerging view that there are unitary neurobiological mechanisms of anesthetic induced sedation and unconsciousness. These observed electroencephalographic differences between N2O and propofol/sevoflurane action might of course be attributable to their differing molecular/cellular sites of action and their pharmacological potency. N2O is reported to predominantly antagonize NMDA receptor activity, whereas propofol and sevoflurane potently potentiate GABAergic receptor activity, as do the benzodiazepines (Rudolph and Antkowiak, 2004). However such a simple explanation is complicated by the fact that xenon, which is also reported to significantly antagonize NMDA receptor activity and to have no GABAergic action, has quite different electroencephalographic properties. Xenon has been demonstrated to slow the EEG and produce an increase in total power especially in frontal regions (Johnson et al., 2003). In particular, xenon has been associated with an increase in slow wave theta activity (Hartmann et al., 1991; Laitio et al., 2008; Utsumi et al., 1998). This suggests contrasting mechanisms of action compared with N2O. An additional consideration is the clear differences in the potency of these agents, with N2O not capable of producing full anesthesia even at high concentrations (>60%), whereas Xenon and the other inductive GABAergic anesthetics show a far greater potency sufficient for surgical anesthesia. Therefore, future research needs to better characterize the dynamical and network level differences between N2O, xenon and propofol/sevoflurane as this will help clarify the plausibility of pursuing unitary mechanistic descriptions of anesthetic action on the basis of macroscopic brain activity. Furthermore, such studies will need to address the inherent limitations of our experimental methodology in which participants were exposed to only one steady-state N2O level. In order to control for inter-subject variability it will be important that subsequent studies involve, where possible, larger sample size repeated measure crossover designs in which multiple agents and concentrations are administered. 5. Financial interests This research was supported by departmental funds to (D.T.J.L.). B.L.F has previously (2009) been paid as a scientific consultant by Cortical Dynamics Ltd., an unlisted subsidiary of Biopharmica Ltd. (Perth, Australia), as part of the development of an EEG based depth of anesthesia monitor. D.T.J.L has an unvalued equity stake in Cortical Dynamics Ltd. D.T.J.L is an inventor on 4 patent applications filed by Cortical Dynamics Ltd. since 2004 that describe new approaches to monitoring depth of anesthesia using the EEG. To date only one patent in one territory (NZ) has been issued. None of the IP declared in any published, pending or granted patent has been licensed. References Alkire MT, Haier RJ, Barker SJ, Shah NK, Wu JC, Kao YJ. Cerebral metabolism during propofol anesthesia in humans studied with positron emission tomography. Anesthesiology 1995;82:393–403. Bennett C, Voss LJ, Barnard JP, Sleigh JW. Practical use of the raw electroencephalogram waveform during general anesthesia: the art and science. Anesth Analg 2009;109:539–50. Boveroux P, Vanhaudenhuyse A, Bruno MA, Noirhomme Q, Lauwick S, Luxen A, et al. Breakdown of within- and between-network resting state functional magnetic resonance imaging connectivity during propofol-induced loss of consciousness. Anesthesiology 2010;113:1038–53.
Breshears JD, Roland JL, Sharma M, Gaona CM, Freudenburg ZV, Tempelhoff R, et al. Stable and dynamic cortical electrophysiology of induction and emergence with propofol anesthesia. Proc Natl Acad Sci USA 2010;107:21170–5. Brown EN, Lydic R, Schiff ND. General anesthesia, sleep, and coma. N Engl J Med 2010;363:2638–50. Brown EN, Purdon PL, Van Dort CJ. General anesthesia and altered states of arousal: a systems neuroscience analysis. Annu Rev Neurosci 2011;34:601–28. Ching S, Cimenser A, Purdon PL, Brown EN, Kopell NJ. Thalamocortical model for a propofol-induced alpha-rhythm associated with loss of consciousness. Proc Natl Acad Sci USA 2010;107:22665–70. Cimenser A, Purdon PL, Pierce ET, Walsh JL, Salazar-Gomez AF, Harrell PG, et al. Tracking brain states under general anesthesia by using global coherence analysis. Proc Natl Acad Sci USA 2011;108:8832–7. Clark MS, Brunick AL. Handbook of nitrous oxide and oxygen sedation. 2nd ed. St Louis, Mo: Mosby; 2003. Delorme A, Makeig S. EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis. J Neurosci Methods 2004;134:9–21. Deshpande G, Kerssens C, Sebel PS, Hu X. Altered local coherence in the default mode network due to sevoflurane anesthesia. Brain Res 2010;1318: 110–21. Deutsch G, Samra SK. Effects of nitrous oxide on global and regional cortical blood flow. Stroke 1990;21:1293–8. Feshchenko VA, Veselis RA, Reinsel RA. Propofol-induced alpha rhythm. Neuropsychobiology 2004;50:257–66. Field LM, Dorrance DE, Krzeminska EK, Barsoum LZ. Effect of nitrous oxide on cerebral blood flow in normal humans. Br J Anaesth 1993;70:154–9. Fitzgibbon SP, Powers DM, Pope KJ, Clark CR. Removal of EEG noise and artifact using blind source separation. J Clin Neurophysiol 2007;24:232–43. Foster BL, Liley DT. Nitrous oxide paradoxically modulates slow electroencephalogram oscillations: implications for anesthesia monitoring. Anesth Analg 2011;113:758–65. Franks NP. General anaesthesia: from molecular targets to neuronal pathways of sleep and arousal. Nat Rev Neurosci 2008;9:370–86. Gugino LD, Chabot RJ, Prichep LS, John ER, Formanek V, Aglio LS. Quantitative EEG changes associated with loss and return of consciousness in healthy adult volunteers anaesthetized with propofol or sevoflurane. Br J Anaesth 2001;87:421–8. Hardmeier M, Zimmermann R, Ruegg S, Pfluger M, Deuster S, Suter K, et al. Intranasal midazolam: pharmacokinetics and pharmacodynamics assessed by quantitative EEG in healthy volunteers. Clin Pharmacol Ther 2012;91:856–62. Hartmann A, Dettmers C, Schuier FJ, Wassmann HD, Schumacher HW. Effect of stable xenon on regional cerebral blood flow and the electroencephalogram in normal volunteers. Stroke 1991;22:182–9. Henrie JR, Parkhouse J, Bickford RG. Alteration of human consciousness by nitrous oxide as assessed electro-encephalography and psychological tests. Anesthesiology 1961;22:247–59. Jensen O, Goel P, Kopell N, Pohja M, Hari R, Ermentrout B. On the human sensorimotor-cortex beta rhythm: sources and modeling. NeuroImage 2005;26:347–55. John ER, Prichep LS, Kox W, Valdes-Sosa P, Bosch-Bayard J, Aubert E, et al. Invariant reversible QEEG effects of anesthetics. Conscious Cogn 2001;10:165–83. Johnson BW, Sleigh JW, Kirk IJ, Williams ML. High-density EEG mapping during general anaesthesia with Xenon and propofol: A pilot study. Anaesth Intensive Care 2003;31:155–63. Jung TP, Makeig S, Humphries C, Lee TW, McKeown MJ, Iragui V, et al. Removing electroencephalographic artifacts by blind source separation. Psychophysiology 2000;37:163–78. Kaisti KK, Langsjo JW, Aalto S, Oikonen V, Sipila H, Teras M, et al. 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. Kaneda Y, Ikuta T, Nakayama H, Kagawa K, Furuta N. Visual evoked potential and electroencephalogram of healthy females during the menstrual cycle. J Med Invest 1997;44:41–6. Knott VJ. Quantitative EEG methods and measures in human psychopharmacological research. Hum Psychopharmacol 2000;15:479–98. Ku SW, Lee U, Noh GJ, Jun IG, Mashour GA. Preferential inhibition of frontal-toparietal feedback connectivity is a neurophysiologic correlate of general anesthesia in surgical patients. PLoS One 2011;6:e25155. Kuizenga K, Kalkman CJ, Hennis PJ. Quantitative electroencephalographic analysis of the biphasic concentration-effect relationship of propofol in surgical patients during extradural analgesia. Br J Anaesth 1998;80:725–32. Kuizenga K, Wierda JM, Kalkman CJ. Biphasic EEG changes in relation to loss of consciousness during induction with thiopental, propofol, etomidate, midazolam or sevoflurane. Br J Anaesth 2001;86:354–60. Laitio RM, Kaskinoro K, Särkelä MOK, Kaisti KK, Salmi E, Maksimow A, et al. Bispectral index, entropy, and quantitative electroencephalogram during single-agent xenon anesthesia. Anesthesiology 2008;108:63–70. Lee U, Muller M, Noh GJ, Choi B, Mashour GA. Dissociable network properties of anesthetic state transitions. Anesthesiology 2011;114:872–81. Lee U, Oh G, Kim S, Noh G, Choi B, Mashour GA. Brain networks maintain a scale-free organization across consciousness, anesthesia, and recovery: evidence for adaptive reconfiguration. Anesthesiology 2010;113:1081–91. Mitchell DJ, McNaughton N, Flanagan D, Kirk IJ. Frontal-midline theta from the perspective of hippocampal ‘‘theta’’. Prog Neurobiol 2008;86:156–85.
B.L. Foster, D.T.J. Liley / Clinical Neurophysiology 124 (2013) 417–423 Murphy M, Bruno MA, Riedner BA, Boveroux P, Noirhomme Q, Landsness EC, et al. Propofol anesthesia and sleep: a high-density EEG study. Sleep 2011;34: 283–91. Pelligrino DA, Miletich DJ, Hoffman WE, Albrecht RF. Nitrous oxide markedly increases cerebral cortical metabolic rate and blood flow in the goat. Anesthesiology 1984;60:405–12. Rampil IJ, Kim JS, Lenhardt R, Negishi C, Sessler DI. Bispectral EEG index during nitrous oxide administration. Anesthesiology 1998;89:671–7. Reinstrup P, Messeter K. Cerebrovascular response to nitrous oxide. Acta Anaesthesiol Scand 1994;38:761–2. Reinstrup P, Ryding E, Algotsson L, Berntman L, Uski T. Effects of nitrous oxide on human regional cerebral blood flow and isolated pial arteries. Anesthesiology 1994;81:396–402. Reinstrup P, Ryding E, Ohlsson T, Sandell A, Erlandsson K, Ljunggren K, et al. Regional cerebral metabolic rate (positron emission tomography) during inhalation of nitrous oxide 50% in humans. Br J Anaesth 2008;100:66–71. Rudolph U, Antkowiak B. Molecular and neuronal substrates for general anaesthetics. Nat Rev Neurosci 2004;5:709–20. Solis-Ortiz S, Ramos J, Arce C, Guevara MA, Corsi-Cabrera M. EEG oscillations during menstrual cycle. Int J Neurosci 1994;76:279–92. Tinker JH, Sharbrough FW, Michenfelder JD. Anterior shift of the dominant EEG rhytham during anesthesia in the Java monkey: correlation with anesthetic potency. Anesthesiology 1977;46:252–9.
423
Utsumi J, Adachi T, Kurata J, Miyazaki Y, Shibata M, Murakawa M, et al. Effect of xenon on central nervous system electrical activity during sevoflurane anaesthesia in cats: Comparison with nitrous oxide. Br J Anaesth 1998;80:628–33. Veselis RA, Reinsel RA, Beattie BJ, Mawlawi OR, Feshchenko VA, DiResta GR, et al. Midazolam changes cerebral blood flow in discrete brain regions: an H2(15)O positron emission tomography study. Anesthesiology 1997;87:1106–17. Whitham EM, Lewis T, Pope KJ, Fitzgibbon SP, Clark CR, Loveless S, et al. Thinking activates EMG in scalp electrical recordings. Clin Neurophysiol 2008;119:1166–75. Whitham EM, Pope KJ, Fitzgibbon SP, Lewis T, Clark CR, Loveless S, et al. Scalp electrical recording during paralysis: quantitative evidence that EEG frequencies above 20 Hz are contaminated by EMG. Clin Neurophysiol 2007;118:1877–88. Williams DJ, Morgan RJ, Sebel PS, Maynard DE. The effect of nitrous oxide on cerebral electrical activity. Anaesthesia 1984;39:422–5. Yamamura T, Fukuda M, Takeya H, Goto Y, Furuwaka K. Fast oscillatory EEG activity by analgesic concentrations of nitrous oxide in man. Anaesth Analg 1981;60:283–8.