Effects of Anesthetics

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Effects of Anesthetics


The exact mechanism by which general anesthetics produce a loss of consciousness (LOC) is complex and incompletely understood. Anesthetics are a diverse group of substances that interact with neural tissue through diverse mechanisms. As the field of anesthetica research moves away from the search for a unitary mechanism of action, and toward the understanding that anesthetics work through multiple mechanisms at multiple sites in the brain and spinal cord, it becomes important to identify those affected sites. Examining the pattern of effects caused by various anesthetics is one method to narrow the search for sites of action. There are a number of reasons why understanding these pattern of activity  is important: it may have clinical implications, it may reflect an agent's expected behavioral effects, it may be a reflection of an agent's molecular mechanism(s) of action and it may reveal insights about the network effects of each agent.

Determining the best agent to use for any particular localized neurologic finding will require a great amount of work, but ultimately it could be of greatt.

This review will discuss the effects of different general anesthetic agents (propofol, isoflurane,) on the brain using different research techniques: EEG and neuroimaging; The effects will be investigated at different levels: from molecular to global brain. Both animal and human studies will be discussed.

General anesthetics

Anesthetics are thought to work by interacting with ion channels that regulate synaptic transmission and membrane potentials in important regions of the brain and spinal cord. These ion-channel targets are differentially sensitive to various anesthetic agents.. As a result anesthetics cause a suppression of cortical activity. This occurs by a direct action of general anesthetics on the cortex, and indirectly by inactivation of the endogenous brain stem and hypothalamic arousal systems (which then cause a secondary cortical shut down). Here, only effects on the cortex will be discussed. The thalamus also changes its mode of activity from a relatively deporalised state, to a hyperpolarized mode that facilitates stereotypical “burst firing” patterns of neuronal activity in corticothalamic networks. This burst-suppression firing pattern can be revealed in the electroencephalogram (EEG). (Alkire and Hudetz).

Unconsciousness or deep anesthesia is associated with highly synchronous δ-frequency (1-4 Hz) activity in thalamic circuits (Voss).

Regional suppressive effects can be visualized by looking at difference images between an on and off anesthetic state. The scrambling effect induced by anesthetics can be visualized with the help of neuroimaging for obtaining information about the pattern of activity induced by specific agents (Alkire, 2005).

Anesthetic agents

Propofol is used to induce and maintain anesthesia. Flumazenil is of benefit in patients who become excessively drowsy after benzodiazepines are used for either diagnostic or therapeutic procedures. is a halogenated ether used for inhalational anesthesia. Together with enflurane and halothane, it replaced the flammable ethers used in the pioneer days of surgery. Its use in human medicine is now starting to decline, being replaced with sevoflurane, desflurane and the intravenous anesthetic propofol. Isoflurane is still frequently used for veterinary anesthesia.

Global effects on the cortex

Global effects on the brain contain cortical suppression via direct cortical actions and blockade of brain stem arousal systems, and/or thalamocortical sensory blockade. (Voss)

Looking at EEG, generally anesthetics initially produce high-frequency oscillations followed by a lower frequency, higher amplitude EEG pattern at or beyond the point at which consciousness is lost, indicating a decrease in neuronal activity. (Franks)

Anesthetics & GABAa receptors

Molecular level

To understand how anesthetics might act on neuronal pathways, it is useful to look at anesthetic affects at the molecular level first. There is growing appreciation that specificity at the molecular level might extend to specificity at the level of neuronal networks.

Anesthetics likely work by interactions with ligand-gated ion channels. These channels both have excitatory and inhibitory influences and they are not expressed uniformly throughout the brain. They tend to cluster in specific areas according to their roles in regulating various excitatory and inhibitory interactions among brain regions. (Alkire, 2005)

It is possible, that increased inhibition within certain brain centers might depress activity in ways that lead to LOC. It has also been suggested that changes in GABAa receptor function (Box 2) could disrupt oscillatory networks or other complex neural interactions essential for conscious perception. (Swindale; Thompson &Wafford)

PET studies

A  few PET studies have been done to investigate the effects of general anesthetics. Some  studies show cerebral anesthetic action on a macroanatomical level, but they cannot provide information why certain brain areas are more or less sensitive to anesthetics.

Using PET ligand technology, comparing results from in vivo and in vitro studies of receptor distribution might provide more information about brain areas affected by anesthetics.

In vitro studies show that anesthetics may act via GABAergic mechanism.

The binding characteristic of C-labelled flumazenil, a specific benzodiazepine antagonist, in the presence of anesthetics allows a direct assessment of anesthetic effects on GABAα receptors in the intact brain. Using this, isoflurane has been shown to increase receptor-specific radioligand binding, dependent on GABA α-receptor density. (Heinke)

This observation provides strong support for the hypothesis that the GABA α -receptor is involved in mediating the action of volatile anesthetics in humans. In addition the radioligand binding measured during anesthesia at 1.5 MAC ( minimum alveolar concentration) was significantly greater than at 1.0 MAC, indicating a dose-related effect of isoflurane on GABAα-receptor ligand-binding.

Alkire et al., correlated data from their in vivo studies with regional distribution patterns of various human receptor binding site densities obtained from ex vivo studies. Propofol's regional cerebral metabolic effects are correlated ( r=-0.86, P<0.0005)  with the regional cerebral distribution of the benzodiazepine binding site densities, which means that the more GABA receptors that are located in an area, the more that area is suppressed during propofol anesthesia.

Regression plot: The regression line shows a linear correlation between the regional metabolic decreases occurring during propofol anesthesia in humans and the known regional benzodiazepine receptor densities. The more benzodiazepine receptors, the more the brain metabolic metabolism will decrease during propofol anesthesia.

Unlike propofol, the regional metabolic suppression observed during isoflurane anesthesia show no correlation with the distribution of benzodiazepine receptor densities. This suggests that this process was not mediated through GABAergic mechanism.

However, Alkire et al, found that an isoflurane-induced suppression in glucose metabolism correlated with muscarinic (acetylcholine) binding density (r=0.85, P=0.03) (fig 2).

Regression plot: There is a linear correlation between the regional metabolic decreases occurring during isoflurane anesthesia in humans and the known regional muscarinic acetylcholine receptor densities.

Brain metabolism was less decreased during isoflurane anesthesia in those regions that had more muscarinic receptors. This seems to indicate that regions with higher muscarinic receptor density are less sensitive to isoflurane, suggesting antagonistic actions of isoflurane and acetylcholine on cerebral metabolism. ( Alkire and Haier 2001; Heinke).

The correlation of isoflurane-induced reductions in glucose metabolism with muscarinic binding density is an interesting finding, as muscarinic signaling in the central nervous system is known to be involved in modulation of consciousness and tend to enhance wakefulness.

Gyulai et al, found that the regional decrease in CMRGlu caused by halothane anesthesia, was not directly linked to GABA α -receptor density.

This finding suggests a mechanism of halothane action other than a pure GABA α ergic one to be responsible for the depression of glucose metabolism. However halothane and the GABAα-receptor agonist muscimol caused a nearly similar metabolic decrease in all regions investigated. This metabolic decrease was significantly enhanced by a combination of both agents. In contrast to the lack of a significant correlation during halothane alone, the changes in regional glucose metabolism induced by either muscimol or muscimol/halothane showed a significant positive correlation with GABA α -receptor density, suggesting a GABA α receptor-associated mechanism in mediating neuronal inhibition by halothane.

In summary, PET shows that propofol seems to act through a GABAergic mechanism. Isoflurane however, seems to have a muscarinic mechanism.

Oscilation patterns


Rhytmic oscillatory activities, whether spontaneous or induced, are commonly recorded in the course of the electroencephalogram (EEG) monitoring, and provide the basis for numerous varied behavior patterns and sensory mechanisms. However there have been few reports on spindling oscillation under deep propofol narcosis.

Direct and indirect inhibitory effects of anesthetic agents on cortical activity are reflected in EEG as a shift from low-amplitude, high-frequency EEG, to high-amplitude, low-frequency activity (indicative of cortical suppression); and the appearance of spindles and K-complexes .

A primary reason for the use of EEG-based monitoring in general anesthesia is to detect and warn the anesthetist that retrievable memories are being formed by the patient.

The scalp measures allow large and slow voltages (generated by large collections of synchronously firing neurons) to be dominant in the EEG signal; as compared to electrocorticogram, which is recorded from electrodes positioned directly into the cortex.

The effects of GABAergic anesthetics to suppress cortical activity are reflected in changes in the EEG, which are more-or-less consistent from one agent to another. At low doses of anesthesia there is a paradoxical increase in amplitude, particularly in the β-frequency range (1-30 Hz). The appearance of this so-called ‘biphasic effect has been related to the development of amnesia in the patient. As the brain concentration of the anesthetic agent increases, the dominant frequency of the EEG slows to the theta (4-8 Hz) and then delta (1-4 Hz) wave bands and its amplitude increases. The transition from high frequency to low frequency activity in the EEG signal reflects thalamocortical hyperpolarisation and synchronized neuronal burst activity.

Animal experiments have shown that during BS-EEG approximately 95% of cortical cells switch over to alternating sequences of phasic depolarizing events ( bursts) and electrical silence (flat periods). While 30-40% of thalamic cells continue to discharge rhythmic spike bursts during flat periods in neocortical neurons.

Wolter et al found that spindle patterns that appear during human sleep (11-15Hz) have nearly the same frequency as the SOs they observed when applying propofol. The ~14 Hz SOs appeared during very deep narcosis mostly within the flat periods of BS-EEG. Motor fields and cingulated areas were particularly important, being repeatedly determined as cortical main spindle sources. Since no movements occurred simultaneously, spindles may play a role in control processes of the behavioral state and of bodily vital functions, which are likewise supported through modules of the motor system. Changes in the features and distribution of oscillations could be related to changing brain stages and to inputs from the periphery. They frequently observed transitions to higher frequencies in close connection with spindle activities.

The differential regional effects evident between agents occur in particular detailed patterns. Understanding these detailed patterns can offer clues to the underlying cellular mechanisms of action for each agent, especially when the agent specific patterns overlap with the specific regional distribution of a known receptor system.

In their study, Ylinen et al. measured rat hippocampal theta activity under urethane. Hippocampal rhythmic slow activity (RSA or theta) is the most studied pattern and has been implicated in several functions, ranging from sensory processing to the voluntary control of movement.

They suggest that a large part of the membrane fluctuation during theta activity is due achloride-mediated GABAα inhibition. In contrast, during hippocampal sharp waves (SPW) events, the synchronous afferent barrage from the CA3 region resulted in a powerful depolarization of the pyramidal cell dendrites. They also showed that volatile anesthetic halothane blocks the mechanism responsible for the emergence of fast oscillation in the CA1 region.

Whole-cell current-clamp recording

Ying et al, performed in an in vitro and in vivo study, a whole-cell current-clamp recording from VB neurons in mouse brain slices.

Propofol decreased the frequency and dampened the regularity of delta oscillations in VB neurons (fig 4). This effect appeared primarily due to the decrease of Ih for the following reasons: propofol at 5 µM inhibited Ih currents; GABAα receptor blockade had little effect on the voltage sag; and 5 µM propofol has a small effect on It. Propofol dampened thalamic oscillations evoked by synaptic stimulation, but this does not indicate a direct effect on glutamatergic responsiveness, because low-concentration propofol (<10 µM) has no effect on glutamatergic transmission.


Synchronous firing of cortical neurons underlies higher forms of neural processing. One possible role of synchrony is that it constitutes a code that signals that disparate low-level features coded by individual neurons in different brain regions belong to the same object (Von de Malsburg). In order for constructing a code, receiving neurons in certain areas must be able to detect the synchrony present among the lower-level feature-detecting neurons. More generally, a fundamental operation of cortical neurons may be to detect the synchronous arrival of sets of impulses from neurons in a number of other cortical areas and to respond with an action potential (Softky).

Swindale (2003) shows in his study that in order for synchrony codes to be precise, transmission time must be independent of path length over all the connected sites between any two cortical areas. Because path lengths vary, developmental mechanisms must compensate for the resulting delay variations. Delay variations could be detected by spike-timing-dependent cues and compensation implemented by systematic changes in axon diameter, myelin thickness, or intermodal distance.

Earlier research suggested that halothane had no effect on conduction velocity in Schaffer collaterals in hippocampal tissue slices at room temperature. Another study showed that a variety of anesthetic agents caused 10% to 20 % increases in conduction velocity in peripheral nerves of human volunteers.

Swindale et al., showed that anesthetics disrupt neural synchrony codes that are essential for higher forms of neural processing, including those aspects that are responsible for consciousness. This disruption occurs as a result of increased conduction velocity in different amounts different types of myelinated axonal fibers of different diameter.

In order for synchronous firing of a,b, and c to be detected, conduction times between the two areas need to vary by less than the integration time window for neurons d and e.


Neurons across the brain are thought to interact with one another, for example forward information between cortical areas and to facilitate the binding of distinct perceptual attributes into a unitary, conscious percept. Disconnections of these functional interactions within neural networks are probably important for drug effects such as amnesia and unconsciousness. Neuroimaging studies indicate that functional connectivity can be examined under anesthesia, and that anesthesia may be associated with changes in network connectivity. Alkire et al demonstrated impaired corticocortical and thalamocortical connectivity at anesthetic concentrations causing unconsciousness. It has also been shown that a decrease of interactivity in motor networks by 0.5 MAC sevoflurane caused a functional dissociation between the two hemispheres. At 1 MAC sevoflurane motor network connectivity was entirely absent. These findings suggest a dose-dependent decrease of synchronized temporal correlations between neurons within functional networks during anesthesia.
Affected regions


Pet study has shown that most anesthetics cause a global suppression in cerebral blood flow (CBF) when consciousness is lost (except for ketamine), although the degree of this suppression is variable. The suppression pattern across the brain is not uniform: certain regions are more deactivated than others. Studies with propofol, sevoflurane and xenon showed deactivation of the thalamus and some midbrain structures that are associated with the ascending reticular activating system, along with varying degrees of deactivation of particular association cortices, such as the precuneus and the posterior cingulated cortex, the cuneus and some localized regions of the frontal cortex (fig 5).

Effects were also seen in the cerebellum, but these varied considerably between anesthetics. Most studies compared the awake state with that following LOC, but a more powerful approach involves measuring changes in CBF as the dose of anesthetic is increased. Work with propofol showed that CBF changes that were induced by vibrotactile stimulation of the forearm were first reduced in the somatosensory cortices, but that LOC only occurred when CBF changes in the thalamus were abolished. A series of imaging studies that measured changes in glucose metabolism also highlighted the importance of thalamic deactivation during anesthetic-induced LOC.



Because axonal conduction and synaptic transmission are relatively unaffected by general anesthetics, the changes in lipid properties must be subtle, and it has never been clear how such effects might cause anesthetisia.

Given this it is proposed here that general anesthetics produce loss of consciousness by differentially changing conduction velocity among different classes of myelinated axons which in turn leads to a disruption of neural synchrony codes essential for the integrated function of cortical areas.

A simple explanation of why anesthetics quickly and reversibly interfere with hogher forms of mental activity, including consciousness, whereas leaving (at low doses) synaptic function and other rate-based forms of neural processing relatively intact.

Other hypothesized mechanisms in particular, those based on interactions with specific receptors such as the gabaa class, are (necessarily) vague as to why these interactions would lead directly to loss of consciousness and do not easily explain the absence of system-specific behavioral neurological signs that are usually associated with receptor-specific drug actions.


Anesthetics eliminate consciousness not only by suppressing all brain activity, but also by stopping neural network interactions.


In vivo effects of anesthetics are mediated at least in part through GABAergic mechanisms. Furthermore, some studies indicate a more complex mechanism for volatile agents in comparison with propofol in producing metabolic suppression during anesthesia. With the expected further development in imaging techniques, PET will be an indispensable tool to get better insights in the molecular mechanism producing anesthesia in vivo.


The fundamental neurophysiological basis of loss of consciousness is commonly (but not always) a suppression of cortical activity. This is reflected as low frequency, synchronous burst-firing activity of cortical neurons (and manifest in the EEG as low frequency oscillations). However the counterexamples that were presented in voss indicate that low-frequency EEG activity is neither necessary for loss of consciousness ( unconsciousness may occur with an activated EEG), nor may it be sufficient for loss of consciousness (an individual may be awake with delta EEG activity, schizophrenia and dementia.



An obvious weakness of the theory is that there is no evidence that anesthetics produce differential changes in conduction velocity in different classes of axon. However there is also no evidence that contradicts this idea, and it should not be difficult to test it experimentally.

The effects on timing may be too small to be significant because they depend on differential changes in conduction velocity. These differences may be too small to have an effect on timing relations that is significant relative to the temporal integration window. This argument cannot be evaluated because the temporal precision of timing relations in the cortex is unknown. However neurons are able to make exquisitely precise submillisecond or even submicrosecond temporal discriminations in some circumstances, and a general level of millisecond-scale precision does not seem unlikely. Of course, if the theory can be shown to be correct by other means, it should be possible to estimate the level of temporal precision required by higher level neural functions by measuring the degree of de-synchrony produced by anesthetics.

…nog meer…

Verder onderzoek: