The new england journal of medicine
review article
From the Department of Anesthesia and
Critical Care, Massachusetts General Hos-
pital (J.A.C., K.W.M., S.A.F.), and the Depart-
ment of Biological Chemistry and Molecu-
lar Pharmacology, Harvard Medical School
(K.W.M.) — both in Boston. Address re-
print requests to Dr. Forman at the Depart-
ment of Anesthesia and Critical Care, CLN-3,
Massachusetts General Hospital, 10 Fruit
St., Boston, MA 02114, or at saforman@
partners.org.
N Engl J Med 2003;348:2110-24.
Copyright © 2003 Massachusetts Medical Society.
2110
Downloaded from www.nejm.org on August 13, 2005 . This article is being provided free of charge for use in Colombia.
Copyright © 2003 Massachusetts Medical Society. All rights reserved.
drug therapy
Alastair J.J. Wood, M.D., Editor
Mechanisms of Actions
of Inhaled Anesthetics
Jason A. Campagna, M.D., Ph.D., Keith W. Miller, D.Phil.,
and Stuart A. Forman, M.D., Ph.D.
uffering so great as i underwent cannot be expressed in
s
“
words . . . but the blank whirlwind of emotion, the horror of great darkness,
and the sense of desertion by God and man, which swept through my mind,
and overwhelmed my heart, I can never forget.”1 Such was the experience of surgery
before October 1846, when William Morton’s successful public demonstration of
ether anesthesia at Massachusetts General Hospital led to its widespread acceptance
by surgeons. Today, anesthesiologists employ a wide variety of drugs, some of which
they use exclusively to produce general anesthesia.2,3 This review focuses on the inhaled
anesthetics (Fig. 1) in current use: nitrous oxide, halothane, enflurane, isoflurane, sevo-
flurane, and desflurane. Our understanding of how these drugs reversibly alter central
nervous system function has changed dramatically in the past two decades.
We will summarize the evidence that inhaled anesthetics ablate movement in re-
sponse to noxious stimuli (immobilization) by depressing spinal cord functions, where-
as their amnesic actions are mediated within the brain. Some compounds, which are
referred to as nonimmobilizers, share many chemical features of inhaled anesthetics
and possess amnesic activity but do not inhibit movement in animals. These differences
suggest that anesthetics induce amnesia and immobility by affecting distinct molecular
targets. Our review will also describe new techniques for dynamically assessing region-
al brain activity, which when combined with electrophysiological and behavioral mon-
itoring, promise to provide important insights into the ways in which anesthetics affect
neural networks. Research on anesthetic-sensitive ion-channel proteins that control
neuronal excitability has revealed that g-aminobutyric acid type A (GABAA) receptors
are most likely involved in the actions of many general anesthetics. Even so, some in-
haled anesthetics may act by inhibiting such excitatory ion channels as neuronal nico-
tinic and glutamate receptors. The subtly different clinical actions of inhaled anesthet-
ics are probably due to distinct, specific actions on a small number of critical molecular
targets.
The identification of specific binding sites for inhaled anesthetics on certain pro-
teins is a dramatic departure from the classic view that all general anesthetics act non-
specifically. Differential sensitivities to various anesthetic actions may have a genetic
basis. Moreover, new research indicates that the effects of general anesthetics depend
on multiple features of their molecular structure and that focusing on such features
may further improve the clinical utility of general anesthetics.4,5
the changing use and role of general anesthesia
Except for the treatment of status epilepticus, general anesthesia is always an adjunct
to another procedure. Anesthetic practice has evolved in response to new procedures,
n engl j med 348;21 www.nejm.org may 22, 2003
 drug therapy

Hydrocarbons Ethers Others

Cl
H H
H C Cl
H3C C O C CH3
Cl N N O
1840¡1950
Chloroform H H Nitrous oxide
CH2 Diethyl ether
H2C CH2
H2C CH2
Cyclopropane Ethylene F F F Cl F F

H C C O C H H C C O C H
F Br
F Cl F Cl F F Xe
1951¡1990
F C C H Enflurane Methoxyflurane Xenon
F H F F H H
F Cl
Halothane F C C O C H F C C O C C

F Cl F F H H H
Isoflurane Fluroxene

CF3 H F F F

1991¡2003 H C O CH2F F C C O C H

CF3 H F H F
Sevoflurane Desflurane

Figure 1. Classes and Generations of Inhaled Anesthetics.

Within a few years after their introduction into widespread clinical use, three major classes of inhaled anesthetics were used: hydrocarbons,
ethers, and other (non–carbon-based) gases. Nitrous oxide was first recognized as an analgesic in the early 19th century, but its low potency
precludes its use as the sole anesthetic agent for most procedures. The hydrocarbons and diethyl ether were either highly toxic (chloroform)
or explosive (cyclopropane, ethylene, and ether). Halogenation of alkanes and ethers reduces their flammability, but fluroxene, the first such
compound introduced in 1954, was later withdrawn from use because of residual combustibility. Halothane, the first noncombustible volatile
halogenated alkane, entered clinical practice in 1956. Enflurane and isoflurane, both halogenated ethers, were first used clinically in 1972 and
1981, respectively. As compared with diethyl ether and halothane, these are less soluble in blood, allowing faster pulmonary uptake and elim-
ination. The uptake and elimination of sevoflurane and desflurane, introduced in the 1990s, are even faster. Xenon, which was first recognized
as an anesthetic in 1951, has highly favorable clinical features including no taste or odor, rapid pulmonary uptake and elimination, no hepatic
or renal metabolism, and minimal cardiovascular depression and arrhythmogenicity. The limited supply of xenon and the expense of extract-
ing it from the atmosphere will most likely prohibit its widespread use in the immediate future.

and in turn, anesthesia has accelerated the devel-
opment of these procedures. The number of ambu-
latory surgical procedures is increasing rapidly in
the United States; nearly 75 percent of all surgical
procedures are now performed on an outpatient ba-
sis.6 General anesthesia is also increasingly used for
noninvasive and minimally invasive diagnostic and
therapeutic techniques that require immobilization
and deep sedation of the patient, as in pediatric ra-
diology and endoscopy, interventional radiology,
electroconvulsive therapy, radiation therapy, vari-
ous cardiologic procedures, transbronchial biopsy,
and urologic procedures. In these settings, which
emphasize cost effectiveness, rapid discharge, and
patient satisfaction, the rapid emergence from an-
esthesia and minimization of side effects are espe-
cially important.
Even though volatile anesthetics can cause car-
diopulmonary depression and death at concen-
trations near those that produce deep anesthesia,
improvements in practice have reduced mortality
attributable to anesthesia to an estimated 1 per
250,000 healthy patients.7 More common undesir-
able and potentially harmful effects that occur dur-
ing and after general anesthesia are autonomic
instability, hypothermia, cardiac dysrhythmias,
nausea, vomiting, and delirium; these effects not
only cause discomfort for the patient but also delay
discharge and increase costs.8 In some cases the
use of general anesthesia outside the operating

n engl j med 348;21 www.nejm.org may 22, 2003 2111

Downloaded from www.nejm.org on August 13, 2005 . This article is being provided free of charge for use in Colombia.
Copyright © 2003 Massachusetts Medical Society. All rights reserved.
 The new england journal
room may pose a greater risk to the patient than the
concomitant procedure itself (e.g., magnetic reso-
nance imaging in children). For these reasons, in-
haled anesthetics that allow rapid emergence of
anesthesia and have few adverse effects are highly
desirable.
what is general anesthesia?
Oliver Wendell Holmes introduced the word “an-
aesthesia” to signify insensibility to surgical pain.
However, there is still no consensus on a more ob-
jective definition of general anesthesia. At different
concentrations inhaled anesthetics induce a variety
of reversible, clinically important effects (Table 1
and Fig. 2). Low concentrations can induce amne-
sia, euphoria, analgesia, hypnosis, excitation, and
hyperreflexia. Higher concentrations cause deep
sedation, muscle relaxation, and diminished motor
and autonomic responses to noxious stimuli, effects
that progress to “surgical” anesthesia. Some volatile
anesthetics also protect the myocardium against
the effects of ischemia, an important component of
anesthetic action for many patients.10
Rigorous definitions have been introduced for
investigations of the underlying mechanisms of
anesthetic effects in humans and animals and for
the clinical assessment of the depth of anesthesia
(Table 1).11 These effects must be reversible and
produced without the need for supplemental mus-
cle relaxants, benzodiazepines, narcotics, or auto-
nomic modulators. Loss of appropriate response
to specific spoken commands is used to identify
hypnosis (the impairment of perceptive awareness)
in anesthetized subjects. In addition, patients may
have perceptive awareness without recall because
memory is more sensitive to anesthetics than aware-
ness.12-14 Laboratory animals are assessed for loss
of righting reflexes — the inability to return to an
upright position in response to nonpainful stimuli
— whereas other tests, such as behavioral condi-
tioning, assess the effects of anesthesia on learn-
ing and memory.
Scales that assess the potency of inhaled anes-
thetics are based on alveolar (in practice, usually
end-expiratory) anesthetic concentrations that are
associated with defined behavioral end points (Ta-
ble 1). The most widely used scale is the minimal
alveolar concentration of anesthetic that suppress-
es purposeful movement in response to a stand-
ard noxious stimulus (MAC or MAC-immobility),
although today the acronym MAC is more frequently
2112 n engl j med 348;21 www.nejm.org
Downloaded from www.nejm.org on August 13, 2005 . This article is being provided free of charge for use in Colombia.
Copyright © 2003 Massachusetts Medical Society. All rights reserved.
of medicine
used to refer to the median alveolar concentration,
indicating the median value for a population under
controlled conditions.9 MAC-immobility increases
as the intensity of the stimulus increases. Analo-
gous potency scales define the MAC that prevents
voluntary responses to spoken commands (MAC-
awake)12,15 and the MAC required to blunt auto-
nomic responses to painful stimuli (MAC-BAR).16
nonspecific pharmacology
and lipid theories
of anesthetic action
For more than a century, two concepts — the unitary
hypothesis and the Meyer–Overton rule — domi-
nated thinking about the mechanisms underlying
anesthetics. By the 1870s, a wide range of structur-
ally unrelated agents were known to possess anes-
thetic activity, leading Claude Bernard to postulate
that all of them acted through a common mecha-
nism.17 Approximately 30 years later, Meyer and
Overton observed a strong correlation between the
potency of anesthetics and their solubility in olive
oil (Fig. 2).18,19 These two ideas led to the theory
that volatile anesthetics act nonspecifically on hy-
drophobic lipid components of cells.
Most researchers have abandoned this theory,
despite its elegance, because anesthetics cause only
slight perturbations in lipids, which can be repro-
duced by small changes in temperature that do not
alter behavior in animals.20 There are, moreover,
a number of apparent exceptions to the Meyer–
Overton correlation. These can be explained by
variations in the size, rigidity, and polarity of the
anesthetic21,22 and by the location of anesthetics
within lipid bilayers, which differs from that of re-
lated compounds without anesthetic activity.23,24
Interest in the possible role of lipids and lipid–pro-
tein interactions in anesthesia continues,22 but
models for the rigorous testing of current hypothe-
ses are lacking. Furthermore, it now appears un-
likely that the different structural classes of inhaled
anesthetics (Fig. 1) act through a single common
mechanism.
behavioral pharmacology
of inhaled anesthetic actions
Behavioral studies have revealed a number of crit-
ical exceptions to the Meyer–Overton rule and the
unitary hypothesis. For instance, homologous series
of anesthetics such as the n-alcohols and n-alkanes
may 22 , 2003
 drug therapy
exhibit a steady increase in potency as successive
methylene (CH2) groups are added, up to a cutoff
point at which adding another methylene elimi-
nates anesthetic activity (the “long-chain alcohol
cutoff ”).25,26 Furthermore, the so-called nonim-
mobilizers, which are volatile halogenated alkanes
with structural similarities to volatile anesthetics,
are predicted by the Meyer–Overton rule to be po-
tent anesthetics. However, they lack immobilizing
activity and in some cases induce convulsions.27 In
contrast to the historical focus on hydrophobicity
alone, analyses of molecular structure and activity
indicate that hydrophobicity, electrostatics, and size
all contribute to the immobilizing potency of in-
haled anesthetics.28-30
Comparing pharmacology among several an-
esthetic actions reveals distinct relations between
structure and activity. For example, the ratios of
MAC-awake to MAC-immobility for nitrous oxide
and diethyl ether are significantly higher than those
for some halogenated volatile anesthetics.12,15 And,
because volatile nonimmobilizers produce amne-
sia in animals,31 it is likely that immobilization and
amnesia are mediated by separate mechanisms.32
Amnesia and hypnosis in humans can also be dis-
tinguished clinically, electrophysiologically, and
pharmacologically.33
anesthetic actions
on different regions
of the nervous system
the spinal cord
It is remarkable that anesthetic-induced ablation of
movement in response to pain is mediated primar-
ily by the spinal cord.34 Experiments in animals
have shown that anesthetic actions in the brain are
not required to inhibit motor responses to pain. In
anesthetized rats, cervical transection of the spinal
cord does not alter the MAC of a given anesthetic
for limb stimulation. Similarly, in goats, selective
administration of isoflurane to the body but not the
brain (the medulla and above) also has little effect
on the concentration that inhibits withdrawal from
hind-leg pain. By contrast, hypnosis and amnesia
are supraspinal effects.32
General anesthetics decrease the transmission
of noxious information ascending from the spinal
cord to the brain, thereby decreasing supraspinal
arousal.35,36 In goats, selective delivery of general
anesthetics to the torso slows cortical electroen-
n engl j med 348;21 www.nejm.org
Downloaded from www.nejm.org on August 13, 2005 . This article is being provided free of charge for use in Colombia.
Copyright © 2003 Massachusetts Medical Society. All rights reserved.
Table 1. Glossary of Terms.
Median alveolar concentration (MAC): The end-tidal concentration (in stand-
ardized pressure units) of inhaled anesthetic that ablates movement (e.g.,
withdrawal) in response to surgical incision in 50 percent of a test popula-
tion. MAC is synonymous with MAC-immobility. The original acronym
MAC stood for minimal alveolar concentration, an end point that meas-
ured the concentration of inhaled anesthetic required to block purposeful
movement in an individual subject.9
Median alveolar concentration awake (MAC-awake): The end-tidal concentra-
tion of inhaled anesthetic that prevents appropriate voluntary responses
to spoken commands (e.g., to open one’s mouth or to raise a hand) in 50
percent of a test population. This end point measures perceptive aware-
ness rather than memory.
Median alveolar concentration for blunting autonomic responses (MAC-BAR):
The end-tidal concentration of inhaled anesthetic that blocks changes in
blood pressure and heart rate in response to surgical incision in 50 per-
cent of a test population.
Loss of righting reflexes (LORR): The failure of an animal to regain upright
posture when placed on its back. Because this end point measures re-
sponses to non-noxious stimuli, its dependence on anesthetic concen-
tration is closely related to MAC-awake, which can only be measured in
humans.
Potency: A measure of relative drug activity that is inversely related to the con-
centration required to produce a standard effect. A volatile anesthetic that
produces a behavioral effect at half the concentration of another anesthet-
ic is said to be twice as potent.
Amnesia: The partial or complete loss of memory. Usually anterograde (affect-
ing recall of experiences after the onset of anesthesia), amnesia may also
be retrograde (affecting recall of experiences that precede the onset of an-
esthesia).
Explicit memory: Information that is consciously perceived and retained and
that can subsequently be reported accurately.
Implicit memory: Information that is unconsciously perceived and retained
and that cannot subsequently be reported. However, this experience af-
fects a subject’s subsequent performance.
Hypnosis: There are various functional definitions of this term. We use it to
connote drug-induced impairment of cognitive functions required for re-
sponding appropriately to environmental stimuli, including attention and
perception. For a patient in the awake state, administration of inhaled an-
esthetics can produce a wide range of hypnotic depths, from mild inatten-
tion to unresponsiveness to noxious stimuli.
Sedation: There are various functional definitions of this term, which is some-
times used as a synonym for “hypnosis.” We use the term to connote drug-
induced hypnosis with anxiolysis, diminished motor activity, and de-
creased arousal.
cephalographic signals.37 Only when volatile anes-
thetics are delivered to the brain in concentrations
that are nearly three times the control MAC do they
produce immobility in goats.38 Therefore, it is like-
ly that ascending signals from the spinal cord af-
fect the hypnotic actions of anesthetics in the brain,
whereas descending signals modify the immobi-
lizing actions of anesthetics in the spinal cord.
may 22, 2003 2113
 The new england journal
the brain
Above the spinal cord, inhaled agents globally de-
press blood flow and glucose metabolism and se-
lectively depress several supraspinal regions.39 For
example, mildly hypnotic concentrations of isoflu-
rane reduce task-induced brain activation in sever-
al distinct cortical regions, whereas activity in the
visual cortex, motor cortex, and subcortical regions
remains unchanged.40 Tomographic assessment
of regional uptake of glucose in deeply anesthetized
volunteers also indicates that the thalamus and
midbrain reticular formation are more depressed
than other regions.41 Evoked potentials traveling
from the periphery to the sensory cortex show in-
creased latency and decreased amplitude in patients
under deep anesthesia with a volatile anesthetic.
This signal degradation is discontinuous, occur-
ring at specific relay sites in the thalamus and the
deep cortex.35
Although there is no definitive evidence that
specific regions of the brain are targets of inhaled
anesthetics, attention has focused on structures with
roles in anesthetic-sensitive functions. The retic-
ular-activating system, thalamus, pons, amygda-
la, and hippocampus are involved in cognition,
memory, learning, sleep, and attentiveness.35,42-45
Interestingly, sleep states and general-anesthesia
states share electroencephalographic and behavior-
al features. In both there is suppression of sensory
input, inhibition of motor output, and analgesia.46
Although sleep and general-anesthesia states are
clearly distinct, subcortical neuronal networks in-
volved in the generation of sleep may also mediate
some anesthetic effects.43,47 Recent studies impli-
cate the tuberomammilary nucleus, a GABA-mod-
ulated region of the hypothalamus that is linked to
sleep states, in the sedative actions of some intrave-
nously administered general anesthetics and per-
haps inhaled agents.48
Most inhaled anesthetics produce generalized
slowing, increased amplitude, and “frontal domi-
nance” of electroencephalographic activity, yet sur-
gical anesthesia has no electroencephalographic
signature. As a result, some measurements derived
from electroencephalographic data correlate well
with hypnotic and immobilizing end points for in-
dividual agents, but no one measure can predict the
depth of anesthesia induced by all inhaled agents
or combinations of these agents.49 Nevertheless,
uncoupling of coherent anteroposterior and inter-
hemispherical electrical activity is consistently as-
sociated with anesthetic-induced unconsciousness
2114 n engl j med 348;21 www.nejm.org
Downloaded from www.nejm.org on August 13, 2005 . This article is being provided free of charge for use in Colombia.
Copyright © 2003 Massachusetts Medical Society. All rights reserved.
of medicine
(in this case, cessation of counting by patients dur-
ing induction).50
molecular actions
of inhaled anesthetics
protein sites
General anesthetics have long been known to inter-
act with small cavities within most globular pro-
teins, but with considerable selectivity.30,51,52 In a
series of seminal experiments, Franks and Lieb es-
tablished that a wide variety of anesthetics inhibit
the lipid-free enzyme firefly luciferase in accord with
the Meyer–Overton rule.53,54 The inhibition of lu-
ciferase even exhibits a long-chain alcohol cutoff,
which is related to the size of the anesthetic-bind-
ing pocket.55 These observations were important
because they demonstrated that protein sites may
also contribute to the effects of general anesthet-
ics.25,56 Although anesthetics alter the functions of
a variety of cytoplasmic signaling proteins, includ-
ing protein kinase C,57,58 the proteins considered
the most likely molecular targets of anesthetics are
ion channels.
effects of anesthetics on ion channels
Ion channels are proteins that regulate the flow of
ions across the cytoplasmic membrane. A variety
of ion channels that modulate the electrical activ-
ity of cells are linked to the behavioral or physio-
logical actions of anesthetics (Table 2). Some of
these channels are sensitive to various inhaled an-
esthetics (Table 3). Ion channels that are sensitive
to volatile anesthetics at clinically effective concen-
trations include both the superfamily of “cysteine-
loop” neurotransmitter receptors, which includes
nicotinic acetylcholine, serotonin type 3, GABAA,
and glycine receptors, and the glutamate receptors
that are activated by N-methyl-d-aspartate (NMDA)
or a-amino-3-hydroxy-5-methyl-4-isoxazolepropi-
onic acid (AMPA).81-84 Within synapses, ion chan-
nels can influence the presynaptic release of neuro-
transmitters and alter postsynaptic excitability in
response to the release of neurotransmitters. Volt-
age-gated ion channels for sodium, potassium, and
calcium are also sensitive to some inhaled anes-
thetics, albeit usually at concentrations higher than
those used clinically.82 A working hypothesis is that
inhaled anesthetics enhance inhibitory postsynap-
tic channel activity (GABAA and glycine receptors)
and inhibit excitatory synaptic channel activity (nic-
otinic acetylcholine, serotonin, and glutamate re-
may 22 , 2003
 drug therapy

ceptors) (Fig. 3). Anesthetic actions at GABAA re- The potency with which volatile anesthetics
ceptors have received the most attention. enhance the function of GABAA receptors in vitro
parallels MAC-immobility.87 Many other classes of
gaba a receptors general anesthetics also enhance GABAA respons-
The GABAA receptors are the most abundant inhib- es,59,88 but nonimmobilizers do not.89 Paralleling
itory neurotransmitter receptors in the brain. Each the enhanced responses of GABAA receptors in
receptor is a heteromeric transmembrane protein vitro, positron-emission tomography in humans
complex that opens a chloride-permeable pore in demonstrates concentration-dependent anesthet-
response to GABA binding (Fig. 3). There are at ic modulation of GABAA receptors in the brain.90
least 18 distinct GABAA-receptor subunit genes in These observations support a central role for GABAA
the human genome, and although most receptor receptors in anesthesia and, until recently, seemed
complexes are thought to contain combinations of to suggest a common mechanism for all inhaled
a, b, and g subunits, a variety of combinations of general anesthetics.
subunits can form functional channels, and the
neuroanatomical distribution of the various types other ion channels
of subunits is not homogeneous.85 At clinically ef- The modulation of GABAA receptors, however, is
fective concentrations, general anesthetics increase neither necessary nor sufficient to account for ev-
the apparent sensitivity of receptors to GABA and
prolong the receptor-mediated inhibitory current
after a pulse of GABA release (Fig. 3). This augments A
GABAA-receptor–mediated inhibition of postsyn- 1.0

aptic neuronal excitability.86

Explicit
0.8 memory
Proportion with No Response

Perceptive
awareness Movement
Figure 2. Potency of Inhaled Anesthetics. in response
The important effects of inhaled anesthetics are pro- to pain
0.6 Autonomic
duced at different concentrations. Panel A illustrates the reactions
general relation between the end-expiratory concentra-
tion of an anesthetic and four anesthetic actions. The ab-
0.4
lation of explicit memory and responsiveness to a
spoken command (hypnosis) are produced at anesthetic
concentrations lower than those that prevent movement
in response to surgical incision, whereas the blockade of 0.2
autonomic responses to pain requires very deep anes-
thesia. Sensitivity to inhaled anesthetics varies from
person to person, and the steepness of the concentra- 0.0
tion–response curves reflects the distribution of individ-
Anesthetic (standard pressure units)
ual responses. Panel B illustrates the strong correlation
between anesthetic potency with respect to a single end
point and the hydrophobicity of the anesthetics shown in B
Figure 1, referred to as the Meyer–Overton rule. The par- Nitrous oxide
1.00
tial pressures of inhaled anesthetics required to prevent Xenon
movement in response to a surgical incision in humans Ethylene
Anesthetic Partial Pressure (atm)

— that is, the median alveolar concentration (MAC), or

MAC-immobility — is plotted against the olive-oil–gas
partition coefficient. Chloroform data are based on stud- 0.10 Cyclopropane
Desflurane
ies in animals. The strong (r=0.99) correlation of MAC-
immobility with anesthetic partitioning between oil and Fluroxene
Diethyl ether
gas phases, a measure of hydrophobicity, can be extend- Sevoflurane Enflurane
ed over a range of potencies of at least 100,000-fold by in- 0.01 Isoflurane
Halothane
cluding other agents that are unsuitable for clinical use
(e.g., nitrogen and thiomethoxyflurane). This correlation Chloroform
indicates that most anesthetics act at one or more hydro-
Methoxyflurane
phobic sites. Although it works for most anesthetics, the 0.00
Meyer–Overton rule fails to explain the lack of anesthetic 1 10 100 1000
potency of some related hydrophobic compounds. Oil¡Gas Partition Coefficient

n engl j med 348;21 www.nejm.org may 22, 2003 2115

Downloaded from www.nejm.org on August 13, 2005 . This article is being provided free of charge for use in Colombia.
Copyright © 2003 Massachusetts Medical Society. All rights reserved.
 The new england journal of medicine

Table 2. Roles of Some Anesthetic-Sensitive Ion Channels in Cellular Excitability, Behavior, Physiological Processes,
and Pharmacology.

Behavioral, Physiological, and

Ion Channel Cellular Roles Pharmacologic Roles

Ligand-gated
g-Aminobutyric acid type A receptors Increased chloride permeability; mem- Enhanced activity associated with anxi-
brane hyperpolarization; inhibition olysis, sedation, amnesia, myorelax-
of excitability ation, anticonvulsant action
Glycine receptors Increased chloride permeability; mem- Spinal reflexes and startle responses;
brane hyperpolarization; inhibition major inhibitory receptor in spinal
of excitability cord
Neuronal nicotinic acetylcholine High permeability to monovalent cat- Association with memory, nociception,
receptors ions and calcium; release of neuro- mutations linked with seizure disor-
transmitters ders; autonomic functions
Muscle nicotinic acetylcholine Neuromuscular transmission Skeletal-muscle contraction
receptors
Serotonin type 3 receptors Enhance excitability by inhibiting rest- Arousal; possible role in emesis
ing potassium-leak currents
Glutamate receptors* Fast excitatory neurotransmission
N-methyl-d-aspartate Cation conductance for calcium and Perception; learning and memory; noci-
magnesium ception
a-Amino-3-hydroxy-5-methyl- Cation conductance for calcium and Perception and memory
4-isoxazole propionic acid magnesium
and kainate
Other types
Potassium channels
Non–voltage-gated background Modulation of cell resting potential Nonspecific role; most likely wide-
channels and excitability; role in chemical, spread
mechanical, and pH sensitivity
Voltage-activated Recovery from action potentials Nerve conduction; cardiac action po-
tentials; mutations associated with
cardiac arrhythmias
Non–voltage-dependent Inward rectifying channels; Glucose sensor in b-cells
neurotransmitter or ATP- pH-sensitive Possible role in ischemic precondition-
activated ing
Sodium channels Generation and propagation of action Nerve conduction; cardiac action po-
potentials tentials (arrhythmias)
Calcium channels
Voltage-gated cardiac (T-, N-, L-, Generation of pacemaker potentials in Cardiac inotropy and chronotropy; vas-
and P-type) neurons (T-type) cular tone
Voltage-gated neuronal Presynaptic localization; neurotrans- Nonspecific role; most likely wide-
mitter release spread
Calcium-induced calcium release
Ryanodine receptor
Inositol triphosphate receptors Intracellular channels Excitation–contraction coupling
Release of intracellular calcium stores
after stimulation of surface recep-
tors; production of calcium oscilla-
tions

* This group excludes metabotropic receptors.

2116 n engl j med 348;21 www.nejm.org may 22 , 2003

Downloaded from www.nejm.org on August 13, 2005 . This article is being provided free of charge for use in Colombia.
Copyright © 2003 Massachusetts Medical Society. All rights reserved.
 drug therapy

Table 3. Functional Effect of Different Classes of Inhaled Anesthetics on Ion Channels.*

Halogenated Xenon and

Alkanes and Nonhalogenated Nitrous
Ion Channel Ethers Alkanes Oxide

g-Aminobutyric acid type A59,60 Enhancement No effect No effect

Glycine receptors60,61 Enhancement No effect No effect
Neuronal nicotinic acetylcholine Strong inhibition Strong inhibition Inhibition
receptors62-65
Muscle nicotinic acetylcholine receptors66 Inhibition Inhibition ND
Serotonin receptors64,67 Weak inhibition ND No effect
Glutamate receptors
N-methyl-d-aspartate64,68,69 Inhibition Inhibition Inhibition
a-Amino-3-hydroxy-5-methyl- Inhibition ND No effect
4-isoxazole propionic acid
and kainate64,69,70
Background potassium channels71,72 Enhancement or ND ND
no effect†
Voltage-activated potassium channels72,73 Inhibition or no ND No effect
effect
ATP-activated potassium channels74 Enhancement or ND ND
no effect†
Voltage-activated sodium channels75,76 Weak inhibition Weak inhibition ND
Voltage-activated calcium channels73,77,78 Weak inhibition ND No effect
Ryanodine-activated calcium channels79,80 Enhancement or ND ND
inhibition

* Effects are indicated qualitatively and represent the preponderance of data at clinical concentrations of anesthetic
agents. In some cases, sensitivities of channels to anesthetics are strongly dependent on the composition of the sub-
units. ND denotes no data.
† The effects are agent-specific.

ery effect of all general anesthetics (Table 3). The
gaseous anesthetics xenon and nitrous oxide only
minimally enhance GABA-mediated currents in
vitro,64,83,91 and even high concentrations of cy-
clopropane and butane fail to alter the function
of GABAA receptors.65 These inhaled anesthet-
ics clearly do not act directly through GABA-medi-
ated mechanisms. Instead, clinical concentrations
of these gases inhibit NMDA-sensitive glutamate
channels and neuronal nicotinic acetylcholine re-
ceptors, suggesting that excitatory ligand-gated ion
channels mediate an alternative pathway to anes-
thesia.
In addition to GABAA receptors, other ion chan-
nels probably have roles in anesthetic-induced im-
mobility. In spinal motor neurons, volatile anes-
thetics augment the activity of inhibitory glycine
receptors69,92 and inhibit postsynaptic AMPA and
NMDA receptors.93 The inhibition of glutamate re-
ceptors is apparently direct and is not due to aug-
mented inhibitory GABA currents.93
Distinct ion channels may mediate different be-
havioral and physiological effects of inhaled anes-
thetics. Neuronal nicotinic acetylcholine receptors
are inhibited by inhaled anesthetics at low concen-
trations that cause amnesia but not immobility, as
well as by the volatile nonimmobilizers.63,94,95 An-
esthetic inhibition of these receptors most likely
impairs memory and learning96 but not immobility.
In the heart, anesthetic inhibition of potassium and
calcium channels is thought to underlie negative
chronotropic and inotropic actions as well as the
pro-arrhythmogenic effects of anesthetics.97-99 The
relative cardiac stability of patients under xenon an-
esthesia as compared with that of patients receiv-
ing halogenated agents correlates with xenon’s

n engl j med 348;21 www.nejm.org may 22, 2003 2117

Downloaded from www.nejm.org on August 13, 2005 . This article is being provided free of charge for use in Colombia.
Copyright © 2003 Massachusetts Medical Society. All rights reserved.
 The new england journal
weaker inhibition of both L-type calcium currents
and voltage-gated potassium currents in cardiac
myocytes.73 The efficacy with which anesthetic
agents produce ischemic preconditioning in the my-
ocardium also correlates with their actions on ATP-
sensitive potassium channels.74
anesthetic sites on ion-channel proteins
Because ion channels function within lipid mem-
branes, it is difficult to discern whether their mod-
ulation by anesthetics is caused indirectly by chang-
es in membrane structure or directly by binding to
protein sites. The most thoroughly verified protein
site is on the peripheral nicotinic acetylcholine re-
ceptor, a structural homologue of both neuronal
nicotinic acetylcholine and GABAA receptors (Fig.
3). Reversible binding of a radio-labeled general
anesthetic was demonstrated with highly purified
peripheral nicotinic acetylcholine receptors.100 The
kinetics of the interruption of the opening of single
nicotinic acetylcholine–receptor channels by anes-
thetics are consistent with the existence of a direct
channel-blocking mechanism but not with an indi-
rect (e.g., lipid-mediated) mechanism.66 Competi-
tion between two anesthetics for an inhibitory site
on open nicotinic acetylcholine–receptor channels
was also demonstrated.101 Experiments involving
mutagenesis, electrophysiology, and photolabel-
ing have mapped the inhibitory site to the nicotinic
acetylcholine–receptor pore.102,103 In the homolo-
gous GABAA receptor, sites that appear critical for
the modulation of volatile anesthetics have also
been identified.104,105 These are located in multiple
transmembrane domains that may form a single
binding pocket (Fig. 3).
integrated models of the
mechanisms of anesthesia
Linking the effects of inhaled anesthetics on specif-
ic ion channels to behavioral effects of general an-
esthesia is a daunting challenge, because the way
neuronal networks influence behavior remains un-
known. Underlying the complexity of these net-
works are their elaborate spatial organization,
the diversity of their synaptic neurotransmitters
and other signaling pathways, and their dynamic
variations in excitability and frequency response to
stimuli. Consequently, the contributions of a spe-
cific class of ion channel to behavior in the animal
as a whole can be extremely difficult to predict. The
actions of inhaled anesthetics have been investigat-
2118 n engl j med 348;21 www.nejm.org
Downloaded from www.nejm.org on August 13, 2005 . This article is being provided free of charge for use in Colombia.
Copyright © 2003 Massachusetts Medical Society. All rights reserved.
of medicine
ed at several network levels, from the simple to the
complex.
synaptic mechanisms
The neuronal functions underlying network activity
are axonal conduction and synaptic transmission.
Clinical concentrations of inhaled anesthetics affect
the latter much more than the former. Inhaled an-
esthetics both depress excitatory synapses and aug-
ment inhibitory synapses.106 Studies aimed at quan-
tifying the presynaptic and postsynaptic effects of
anesthetics have demonstrated actions on both the
release of neurotransmitters and the function of
neurotransmitter receptors, with the latter having a
dominant role.106 In addition, some volatile anes-
Figure 3 (facing page). Anesthetic Sites on Ligand-Gated
Ion Channels.
Two structurally related, but functionally opposing
postsynaptic channels respond in opposite ways to anes-
thetics and appear to possess anesthetic binding sites
located in different regions. The central drawing depicts
structural features of the homologous g-aminobutyric
acid type A (GABAA) receptor and nicotinic acetylcholine
receptor. Each receptor contains five subunits (the front
one is removed for clarity) that cross the lipid bilayer and
are arranged around a central ion channel. Each subunit
is thought to have four transmembrane elements, as in-
dicated in the left-most subunit. Agonist binding sites
are thought to be formed at subunit interfaces in the ex-
tracellular portions of the receptors. In the upper right-
hand panel, nicotinic acetylcholine receptors are excitatory
channels permeable to cations. Their activation depolar-
izes neurons and muscles, making the generation of ac-
tion potentials more likely. In the lower right-hand panel,
excitatory postsynaptic currents activated by acetylcho-
line are noncompetitively inhibited by anesthetics, and
in single-channel currents, anesthetics cause frequent
closures or block the opening of currents. Mutations that
alter the hydrophobicity of the pore-forming regions of
muscle nicotinic acetylcholine receptors alter the sensi-
tivity to inhibition by anesthetics, and photoactivatable
anesthetics covalently incorporate in this region. In the
upper left-hand panel, GABAA receptors are inhibitory
channels permeable to chloride anions. Their activation
hyperpolarizes neurons, making the generation of action
potentials less likely. In the lower left-hand panel, in-
hibitory postsynaptic currents activated by GABA are
prolonged by anesthetics, resulting in supranormal chlo-
ride influx and reduced excitability. GABA concentration–
response curves are shifted leftward, so that low concen-
trations of GABA in cerebrospinal fluid may produce a
chloride leak by means of extrasynaptic GABAA recep-
tors, suppressing neuronal excitation. Studies involving
chimeras and amino acid mutations suggest that anes-
thetics interact at GABAA-receptor sites formed between
several transmembrane elements.
may 22, 2003
 drug therapy

Inhibitory synapse Excitatory synapse

Presynaptic Presynaptic
neuron Action neuron Action
potential potential

Release of
Ca2+ acetylcholine Ca2+

GABA
release

K+ K+ K+
K+
Cl- Na+
Cl- Na+
Cl- Cl- Na+
Cl- Na+ Na+ Na+
Cl-
Postsynaptic Cl- Cl- Postsynaptic Na+ Na+
membrane Cl- Cl- membrane Na
+ Na+ Na+

Agonist binding sites

Anesthetic Anesthetic site on

site on nicotinic acetylcholine
GABAA receptor
receptor

Inhibitory Excitatory
Postsynaptic Postsynaptic
GABAergic Currents Cholinergic Currents
No With
anesthetic anesthetic
No With
anesthetic anesthetic 2
milliseconds
0.1
second
Membrane Current

Membrane Current

With
anesthetic No
Transmembrane anesthetic
element
No Pore
anesthetic With
anesthetic

Log GABA Log Acetylcholine

thetics cause hyperpolarization and diminished recording in animals often requires sedating them
excitability of neurons by enhancing the activity of with other anesthetic agents.109 Many in vitro stud-
background potassium channels.107,108 ies have recorded electrical activity in brain or spinal
cord slices, which maintain local synaptic interac-
in vivo and in vitro neural networks tions. Small cortical slices with intact local networks
Elucidating the effects of inhaled anesthetics on demonstrate synchronized electrical rhythms that
neural networks in vivo has proved technically dif- can be slowed either by enhancing GABA-mediat-
ficult, in part because reliable electrophysiological ed transmission or by inhibiting glutamate-medi-

n engl j med 348;21 www.nejm.org may 22, 2003 2119

Downloaded from www.nejm.org on August 13, 2005 . This article is being provided free of charge for use in Colombia.
Copyright © 2003 Massachusetts Medical Society. All rights reserved.
 The new england journal
ated transmission with general anesthetics.110 The
effects of anesthetics on the frequencies of local
cortical networks probably contribute directly to
electroencephalographic slowing. On the other
hand, long-range (e.g., thalamocortical) circuits
coordinate rhythmic activity among distant brain
regions.41 The effects of anesthetics on the rhyth-
mic frequency of these networks may depend on
the rates of decay of GABAA receptor–mediated in-
hibitory potentials,111 which are prolonged by in-
haled anesthetics (Table 3). The hippocampus and
spinal cord also contain circuits that are most likely
involved in the amnesic and immobilizing actions
of general anesthetics.69,110
Highly simplified neural circuits have proved
more amenable to analysis. The effects of inhaled
anesthetics on respiration have been investigated
by recording caudal ventral respiratory neurons of
the medulla in decerebrate dogs.112,113 Clinically
effective concentrations of inhaled anesthetics alter
both glutamate-mediated and GABA-mediated sig-
nals to pacing neurons; sevoflurane reduces output
motor-neuron activity more than does halothane,112
paralleling the clinical observation that sevoflurane
depresses respiration more deeply than does hal-
othane.114
genetic studies of inhaled
anesthetic actions
Genetic manipulation of animals is another tech-
nique for investigating links between potential tar-
gets of anesthetics and the behavioral effects of
these agents. Two approaches have been used:
screening for mutations in selectively bred popula-
tions and the creation of mutants at putative target
sites.115
genetic screening
Selective breeding and large-scale screening for
mutations in fruit flies (Drosophila melanogaster)
and nematodes (Caenorhabditis elegans) have identi-
fied strains with altered sensitivities to anesthet-
ics,116-120 but the applicability of these data to hu-
mans is questionable, because some behavioral end
points required anesthetic concentrations that are
much greater than the MAC values in mammals. Se-
lective breeding and screening of anesthesia-resist-
ant murine populations, although promising, have
so far failed to identify specific target genes.121-123
2120 n engl j med 348;21 www.nejm.org
Downloaded from www.nejm.org on August 13, 2005 . This article is being provided free of charge for use in Colombia.
Copyright © 2003 Massachusetts Medical Society. All rights reserved.
of medicine
studies in genetically modified mammals
Knockout Studies
Animals in which specific genes have been knocked
out have been used to evaluate the role of two lig-
and-gated ion channels in anesthesia, the AMPA-
sensitive glutamate receptor and the GABAA recep-
tor. The GluR2 subunit of AMPA receptors was
chosen for study, because it is the most common
subunit and it determines anesthetic sensitivity in
vitro.93,124 In GluR2-knockout mice, the MAC of
volatile agents was unaltered, sensitivity to loss of
righting reflexes was moderately increased, and no-
ciception was increased.125 The genes for the b3
and a6 subunits of GABAA receptors were targeted,
because of their patterns of expression in the brain
and the anesthetic sensitivity that they confer on in
vitro receptors containing different types of sub-
units.126 In GABAA b3–knockout mice, the MAC of
enflurane was slightly diminished, but there was no
change in sensitivity to the loss of righting reflex-
es.127-129 In GABAA a6–knockout animals, there
was no change in the MAC of volatile anesthetics or
in the sensitivity to the loss of righting reflexes.130
Several themes emerged from studies of such
knockout mice. First, anesthetic sensitivity meas-
ured on the basis of the loss of righting reflexes and
MAC-immobility were affected in different ways by
genetic alteration, adding to the evidence that dif-
ferent anesthetic-induced forms of behavior are me-
diated by distinct mechanisms. Second, the mech-
anisms that mediate even a single end point, such
as immobility, are complex and agent-dependent,
as shown by the following: knockout of the GABAA
b3 gene decreased the MAC of enflurane but had
far less effect on the MAC of halothane and no ef-
fect on the enflurane-induced depression of evoked
spinal motor potentials.127,129 The data on GABAA-
receptor–knockout mice also suggest that specific
types of receptor subunits, specifically b3, may have
dominant roles in some anesthetic actions. Other
mice have been created in which GABAA-receptor
a1, b2, and d subunits have been inactivated, but
the sensitivity of these animals to inhaled anesthet-
ics has not yet been reported.
“Knock-In” Studies
Knocking out the expression of an ion-channel–
subunit gene may induce changes in the composi-
tion of the subunits, the network circuitry, or
both.131 The introduction of specific mutations into
may 22, 2003
 drug therapy

native genes (“knock-in” animals) avoids these pit-on nonspecific biophysical mechanisms. Reevalu-
falls and enables an assessment of the physiologic ation of the actions of anesthetics on a variety of
and pharmacologic roles of specific proteins and neurobiologic-systems levels has revealed impor-
even small regions within proteins. The power of tant new insights into mechanisms that contradict
this approach is evident from recent revelations these nonspecific hypotheses. Thus, although all
about the roles of specific GABAA-receptor subunitsinhaled general anesthetics produce amnesia and
on the actions of benzodiazepines, which induce suppress motor responses to noxious stimuli, their
some behavioral effects similar to those of generalactions on other behavioral and physiological re-
anesthetics.132 Knock-in models are also being em- sponses vary. The suppression of nociceptive motor
ployed to examine the role of GABAA receptors in responses by anesthetics is mediated primarily by
mediating general anesthesia. A mutation in the the spinal cord, whereas hypnosis and amnesia are
GABAA receptor b3 subunit that attenuates in vitro mediated within the brain. These actions may also
modulation by the intravenous anesthetic etomi- be associated with separate molecular targets. Im-
date has been introduced into mice. These animals portant actions of inhaled anesthetics are associat-
have dramatically reduced sensitivity to etomidate ed with altered activity of neuronal ion channels,
with respect to the end points of nociceptive with-particularly the fast synaptic neurotransmitter re-
drawal and the loss of righting reflexes, but theirceptors such as nicotinic acetylcholine, GABAA, and
sensitivities to volatile anesthetics are only moder-
glutamate receptors. There is also growing evi-
ately decreased (with respect to MAC) or unaffecteddence that anesthetics affect neuronal ion channels
(with respect to the loss of righting reflexes).133 This
by binding directly to protein sites. Different ion
result is further evidence of a role for the b3 sub-
channels display strikingly unique sensitivities to
unit in determining MAC and is consistent with in various inhaled anesthetics, suggesting that differ-
vitro studies showing that mutations in a subunits ent channels are involved in distinct behavioral ef-
have a greater effect on the sensitivity of GABAA fects of anesthetics and that several mechanistic
receptors to volatile anesthetics than do mutationspathways may converge to produce similar anes-
in b subunits.104 Studies using knock-in animals thetic states. Neuroanatomical differences in the
with these a-subunit mutations are expected to pro-distribution of various ion channels and their spe-
vide additional insights into the roles of GABAA re-
cific subunits are likely to influence specific behav-
ceptors in the actions of inhaled anesthetics. ioral effects of inhaled anesthetics. This emerging
view of the specific neurobiologic actions of inhaled
summary anesthetics suggests that these widely used drugs
should be amenable to improvements by rational
Simplifying assumptions such as the unitary hy- design.
pothesis led early research on anesthesia to focus

references
1. Ashhurst J Jr. Surgery before the days of
anaesthesia. In: Warren JC, White JC, Rich-
ardson WL, Beach HH, Shattuck FC, Bigelow
WS, eds. The semi-centennial of anaesthe-
sia, October 16, 1846–October 16, 1896. Bos-
ton: Massachusetts General Hospital, 1897:
27-37.
2. Wiklund RA, Rosenbaum SH. Anesthe-
siology. N Engl J Med 1997;337:1132-41,
1215-9.
3. Urban BW, Bleckwenn M. Concepts and
correlations relevant to general anaesthesia.
Br J Anaesth 2002;89:3-16.
4. McLeod HL, Evans WE. Pharmacoge-
nomics: unlocking the human genome for
better drug therapy. Annu Rev Pharmacol
Toxicol 2001;41:101-21.
5. Weinstein JN. Pharmacogenomics —
teaching old drugs new tricks. N Engl J Med
2000;343:1408-9.
6. Twersky RS. Ambulatory surgery update.
Can J Anaesth 1998;45:Suppl:R76-R83.
7. Kohn LT, Corrigan JM, Donaldson MS,
eds. To err is human: building a safer
health system. Washington, D.C.: National
Academy Press, 2000.
8. Carroll NV, Miederhoff PA, Cox FM,
Hirsch JD. Costs incurred by outpatient sur-
gical centers in managing postoperative
nausea and vomiting. J Clin Anesth 1994;6:
364-9.
9. Eger EI II, Saidman LJ, Brandstater B.
Minimum alveolar anesthetic concentration:
a standard of anesthetic potency. Anesthesi-
ology 1965;26:756-63.
10. Cope DK, Impastato WK, Cohen MV,
Downey JM. Volatile anesthetics protect the
ischemic rabbit myocardium from infarc-
tion. Anesthesiology 1997;86:699-709.
11. Antognini JF, Carstens E, Raines DE,
eds. Neural mechanisms of anesthesia. Con-
temporary clinical neuroscience. Totowa,
N.J.: Humana Press, 2003:3-10.
12. Dwyer R, Bennett HL, Eger EI II, Heil-
bron D. Effects of isoflurane and nitrous
oxide in subanesthetic concentrations on
memory and responsiveness in volunteers.
Anesthesiology 1992;77:888-98.
13. Dwyer R, Bennett HL, Eger EI II, Peter-
son N. Isoflurane anesthesia prevents uncon-
scious learning. Anesth Analg 1992;75:107-
12.
14. Ghoneim MM, Block RI. Learning and
memory during general anesthesia: an
update. Anesthesiology 1997;87:387-410.
15. Stoelting RK, Longnecker DE, Eger EI
II. Minimum alveolar concentrations in man
on awakening from methoxyflurane, halo-
thane, ether and fluroxene anesthesia: MAC
awake. Anesthesiology 1970;33:5-9.

n engl j med 348;21 www.nejm.org may 22, 2003 2121

Downloaded from www.nejm.org on August 13, 2005 . This article is being provided free of charge for use in Colombia.
Copyright © 2003 Massachusetts Medical Society. All rights reserved.
 The
16. Roizen MF, Horrigan RW, Frazer BM.
Anesthetic doses blocking adrenergic (stress)
and cardiovascular responses to incision —
MAC BAR. Anesthesiology 1981;54:390-8.
17. Leake CD. Claude Bernard and anesthe-
sia. Anesthesiology 1971;35:112-3.
18. Overton E. Studien über die Narkose
Zugleich ein Beitrag zur Allgemeinen Phar-
makologie. Jena, Germany: Verlag von
Gustav Fisher, 1901.
19. Meyer H. Zur Theorie der Alkoholnar-
kose. Arch Exp Pathol Pharmakol 1899;42:
109-18.
20. Franks NP, Lieb WR. Molecular mecha-
nisms of general anaesthesia. Nature 1982;
300:487-93.
21. Pohorille A, Cieplak P, Wilson MA. Inter-
actions of anesthetics with the membrane-
water interface. Chem Phys 1996;204:337-
45.
22. Cantor RS. Breaking the Meyer-Over-
ton rule: predicted effects of varying stiff-
ness and interfacial activity on the intrinsic
potency of anesthetics. Biophys J 2001;80:
2284-97.
23. North C, Cafiso DS. Contrasting mem-
brane localization and behavior of halogen-
ated cyclobutanes that follow or violate the
Meyer-Overton hypothesis of general anes-
thetic potency. Biophys J 1997;72:1754-61.
24. Tang P, Yan B, Xu Y. Different distribu-
tion of fluorinated anesthetics and nonanes-
thetics in model membrane: a 19F NMR
study. Biophys J 1997;72:1676-82.
25. Franks NP, Lieb WR. Where do general
anesthetics act? Nature 1978;274:339-42.
26. Raines DE, Korten SE, Hill AG, Miller
KW. Anesthetic cutoff in cycloalkanemetha-
nols: a test of current theories. Anesthesiol-
ogy 1993;78:918-27.
27. Koblin DD, Chortkoff BS, Laster MJ,
Eger EI II, Halsey MJ, Ionescu P. Polyhaloge-
nated and perfluorinated compounds that
disobey the Meyer-Overton hypothesis.
Anesth Analg 1994;79:1043-8.
28. Abraham MH, Lieb WR, Franks NP. Role
of hydrogen bonding in general anesthesia.
J Pharm Sci 1991;80:719-24.
29. Sewell JC, Halsey MJ. Molecular similar-
ity analysis: an alternative approach to study-
ing molecular mechanisms of anaesthesia.
Toxicol Lett 1998;100-101:359-64.
30. Miller KW. The nature of sites of general
anaesthetic action. Br J Anaesth 2002;89:
17-31.
31. Kandel L, Chortkoff BS, Sonner J, Laster
MJ, Eger EI II. Nonanesthetics can suppress
learning. Anesth Analg 1996;82:321-6.
32. Eger EI II, Koblin DD, Harris RA, et al.
Hypothesis: inhaled anesthetics produce
immobility and amnesia by different mecha-
nisms at different sites. Anesth Analg 1997;
84:915-8.
33. Veselis RA, Reinsel RA, Feshchenko VA.
Drug-induced amnesia is a separate phe-
nomenon from sedation: electrophysiologic
evidence. Anesthesiology 2001;95:896-907.
34. Antognini JF, Carstens E. In vivo charac-
2122 n engl j med 348;21
Downloaded from www.nejm.org on August 13, 2005 . This article is being provided free of charge for use in Colombia.
Copyright © 2003 Massachusetts Medical Society. All rights reserved.
new england journal of medicine
terization of clinical anaesthesia and its
components. Br J Anaesth 2002;89:156-66.
35. Angel A. Central neuronal pathways and
the process of anaesthesia. Br J Anaesth
1993;71:148-63.
36. Collins JG, Kendig JJ, Mason P. Anes-
thetic actions within the spinal cord: contri-
butions to the state of general anesthesia.
Trends Neurosci 1995;18:549-53.
37. Antognini JF, Wang XW, Carstens E. Iso-
flurane action in the spinal cord blunts elec-
troencephalographic and thalamic-reticular
formation responses to noxious stimulation
in goats. Anesthesiology 2000;92:559-66.
38. Antognini JF, Schwartz K. Exaggerated
anesthetic requirements in the preferentially
anesthetized brain. Anesthesiology 1993;
79:1244-9.
39. Heinke W, Schwarzbauer C. In vivo
imaging of anaesthetic action in humans:
approaches with positron emission tomog-
raphy (PET) and functional magnetic reso-
nance imaging (fMRI). Br J Anaesth 2002;
89:112-22.
40. Idem. Subanesthetic isoflurane affects
task-induced brain activation in a highly spe-
cific manner: a functional magnetic reso-
nance imaging study. Anesthesiology 2001;
94:973-81.
41. Alkire MT, Haier RJ, Fallon JH. Toward a
unified theory of narcosis: brain imaging
evidence for a thalamocortical switch as the
neurophysiologic basis of anesthetic-induced
unconsciousness. Conscious Cogn 2000;9:
370-86.
42. Hobson JA, McCarley RW, Pivik RT,
Freedman R. Selective firing by cat pontine
brain stem neurons in desynchronized sleep.
J Neurophysiol 1974;37:497-511.
43. Goodman SJ, Mann PE. Reticular and
thalamic multiple unit activity during wake-
fulness, sleep and anesthesia. Exp Neurol
1967;19:11-24.
44. Kendig JJ, MacIver MB, Roth SH. Anes-
thetic actions in the hippocampal formation.
Ann N Y Acad Sci 1991;625:37-53.
45. Wakasugi M, Hirota K, Roth SH, Ito Y.
The effects of general anesthetics on excita-
tory and inhibitory synaptic transmission in
areas CA1 of the rat hippocampus in vitro.
Anesth Analg 1999;88:676-80.
46. Lydic R, Baghdoyan HA. Cholinergic
contributions to the control of conscious-
ness. In: Yaksh TL, Lynch C III, Zapol WM,
Maze M, Biebuyck JF, Saidman LJ, eds. Anes-
thesia: biologic foundations. Philadelphia:
Lippincott–Raven, 1997:433-50.
47. Keifer JC, Baghdoyan HA, Lydic R. Pon-
tine cholinergic mechanisms modulate the
cortical electroencephalographic spindles of
halothane anesthesia. Anesthesiology 1996;
84:945-54.
48. Nelson LE, Guo TZ, Lu J, Saper CB,
Franks NP, Maze M. The sedative component
of anesthesia is mediated by GABA(A) recep-
tors in an endogenous sleep pathway. Nat
Neurosci 2002;5:979-84.
49. Drummond JC. Monitoring depth of
www.nejm.org may 22, 2003
anesthesia: with emphasis on the applica-
tion of the bispectral index and the middle
latency auditory evoked response to the pre-
vention of recall. Anesthesiology 2000;93:
876-82.
50. John ER, Prichep LS, Kox W, et al. Invari-
ant reversible QEEG effects of anesthetics.
Conscious Cogn 2001;10:165-83. [Erratum,
Conscious Cogn 2002;11:138.]
51. Franks NP, Jenkins A, Conti E, Lieb WR,
Brick P. Structural basis for the inhibition of
firefly luciferase by a general anesthetic.
Biophys J 1998;75:2205-11.
52. Eckenhoff RG. Promiscuous ligands
and attractive cavities: how do the inhaled
anesthetics work? Mol Interventions 2001;
1:258-68.
53. Middleton AJ, Smith EB. General anaes-
thetics and bacterial luminescence. I. The
effect of diethyl ether on the in vivo light
emission of Vibrio fischeri. Proc R Soc Lond
B Biol Sci 1975;193:159-71.
54. Franks NP, Lieb WR. Do general anaes-
thetics act by competitive binding to specific
receptors? Nature 1984;310:599-601.
55. Idem. Mapping of general anesthetic tar-
get sites provides a molecular basis for cut-
off effects. Nature 1985;316:349-51.
56. Miller KW. The nature of the site of gen-
eral anesthesia. Int Rev Neurobiol 1985;27:
1-61.
57. Slater SJ, Cox KJ, Lombardi JV, et al. Inhi-
bition of protein kinase C by alcohols and
anaesthetics. Nature 1993;364:82-4.
58. Hemmings HC Jr, Adamo AI. Effects of
halothane and propofol on purified brain
protein kinase C activation. Anesthesiology
1994;81:147-55.
59. Olsen RW. The molecular mechanism
of action of general anesthetics: structural
aspects of interactions with GABAA recep-
tors. Toxicol Lett 1998;100-101:193-201.
60. Harrison NL, Kugler JL, Jones MV,
Greenblatt EP, Pritchett DB. Positive modu-
lation of human gamma-aminobutyric acid
type A and glycine receptors by the inhala-
tion anesthetic isoflurane. Mol Pharmacol
1993;44:628-32.
61. Mascia MP, Machu TK, Harris RA.
Enhancement of homomeric glycine recep-
tor function by long-chain alcohols and
anaesthetics. Br J Pharmacol 1996;119:
1331-6.
62. Flood P, Role LW. Neuronal nicotinic
acetylcholine receptor modulation by gen-
eral anesthetics. Toxicol Lett 1998;100-101:
149-53.
63. Violet JM, Downie DL, Nakisa RC, Lieb
WR, Franks NP. Differential sensitivities of
mammalian neuronal and muscle nicotinic
acetylcholine receptors to general anesthet-
ics. Anesthesiology 1997;86:866-74.
64. Yamakura T, Harris RA. Effects of gas-
eous anesthetics nitrous oxide and xenon on
ligand-gated ion channels: comparison with
isoflurane and ethanol. Anesthesiology 2000;
93:1095-101.
65. Raines DE, Claycomb RJ, Scheller M,

Forman SA. Nonhalogenated alkane anes-
thetics fail to potentiate agonist actions on
two ligand-gated ion channels. Anesthesiol-
ogy 2001;95:470-7.
66. Dilger JP, Vidal AM, Mody HI, Liu Y. Evi-
dence for direct actions of general anesthet-
ics on an ion channel protein: a new look at a
unified mechanism of action. Anesthesiol-
ogy 1994;81:431-42.
67. Jenkins A, Franks NP, Lieb WR. Actions
of general anaesthetics on 5-HT3 receptors
in N1E-115 neuroblastoma cells. Br J Phar-
macol 1996;117:1507-15.
68. Perouansky M, Baranov D, Salman M,
Yaari Y. Effects of halothane on glutamate
receptor-mediated excitatory postsynaptic
currents: a patch-clamp study in adult mouse
hippocampal slices. Anesthesiology 1995;
83:109-19.
69. Kendig JJ. In vitro networks: subcortical
mechanisms of anaesthetic action. Br J
Anaesth 2002;89:91-101.
70. Lin LH, Chen LL, Harris RA. Enflurane
inhibits NMDA, AMPA, and kainate-induced
currents in Xenopus oocytes expressing
mouse and human brain mRNA. FASEB J
1993;7:479-85.
71. Patel AJ, Honore E, Lesage F, Fink M,
Romey G, Lazdunski M. Inhalational anes-
thetics activate two-pore-domain back-
ground K+ channels. Nat Neurosci 1999;2:
422-6.
72. Yost CS. Potassium channels: basic
aspects, functional roles, and medical sig-
nificance. Anesthesiology 1999;90:1186-
203.
73. Huneke R, Jungling E, Skasa M, Ros-
saint R, Luckhoff A. Effects of the anesthetic
gases xenon, halothane, and isoflurane on
calcium and potassium currents in human
atrial cardiomyocytes. Anesthesiology 2001;
95:999-1006.
74. Kohro S, Hogan QH, Nakae Y, Yamakage
M, Bosnjak ZJ. Anesthetic effects on mito-
chondrial ATP-sensitive K channel. Anes-
thesiology 2001;95:1435-40.
75. Lingamaneni R, Birch ML, Hemmings
HC Jr. Widespread inhibition of sodium
channel-dependent glutamate release from
isolated nerve terminals by isoflurane and
propofol. Anesthesiology 2001;95:1460-6.
[Erratum, Anesthesiology 2002;96:782.]
76. Rehberg B, Xiao YH, Duch DS. Central
nervous system sodium channels are signif-
icantly suppressed at clinical concentrations
of volatile anesthetics. Anesthesiology 1996;
84:1223-33.
77. Bosnjak ZJ, Aggarwal A, Turner LA,
Kampine JM, Kampine JP. Differential effects
of halothane, enflurane, and isoflurane on
Ca2+ transients and papillary muscle tension
in guinea pigs. Anesthesiology 1992;76:123-
31.
78. Bleakman D, Jones MV, Harrison NL.
The effects of four general anesthetics on
intracellular [Ca2+] in cultured rat hippo-
campal neurons. Neuropharmacology 1995;
34:541-51.
n engl j med 348;21
Downloaded from www.nejm.org on August 13, 2005 . This article is being provided free of charge for use in Colombia.
Copyright © 2003 Massachusetts Medical Society. All rights reserved.
drug therapy
79. Lynch C III, Frazer MJ. Anesthetic alter-
ation of ryanodine binding by cardiac cal-
cium release channels. Biochim Biophys Acta
1994;1194:109-17.
80. Frazer MJ, Lynch C III. Halothane and
isoflurane effects on Ca2+ fluxes of isolated
myocardial sarcoplasmic reticulum. Anes-
thesiology 1992;77:316-23.
81. Franks NP, Lieb WR. Which molecular
targets are most relevant to general anaes-
thesia? Toxicol Lett 1998;100-101:1-8.
82. Idem. Molecular and cellular mechanisms
of general anaesthesia. Nature 1994;367:
607-14.
83. Mennerick S, Jevtovic-Todorovic V,
Todorovic SM, Shen W, Olney JW, Zorumski
CF. Effect of nitrous oxide on excitatory and
inhibitory synaptic transmission in hippo-
campal cultures. J Neurosci 1998;18:9716-
26.
84. Narahashi T, Aistrup GL, Lindstrom JM,
et al. Ion channel modulation as the basis
for general anesthesia. Toxicol Lett 1998;
100-101:185-91.
85. Macdonald RL, Olsen RW. GABAA recep-
tor channels. Annu Rev Neurosci 1994;17:
569-602.
86. Jones MV, Harrison NL. Effects of vola-
tile anesthetics on the kinetics of inhibitory
postsynaptic currents in cultured rat hippo-
campal neurons. J Neurophysiol 1993;70:
1339-49.
87. Zimmerman SA, Jones MV, Harrison NL.
Potentiation of gamma-aminobutyric acidA
receptor CI– current correlates with in vivo
anesthetic potency. J Pharmacol Exp Ther
1994;270:987-91.
88. Harris RA, Mihic SJ, Brozowski S, Had-
ingham K, Whiting PJ. Ethanol, flunitraz-
epam, and pentobarbital modulation of
GABAA receptors expressed in mammalian
cells and Xenopus oocytes. Alcohol Clin Exp
Res 1997;21:444-51.
89. Mihic SJ, McQuilkin SJ, Eger EI II,
Ionescu P, Harris RA. Potentiation of gamma-
aminobutyric acid type A receptor-mediated
chloride currents by novel halogenated com-
pounds correlates with their abilities to
induce general anesthesia. Mol Pharmacol
1994;46:851-7.
90. Gyulai FE, Mintun MA, Firestone LL.
Dose-dependent enhancement of in vivo
GABA(A)-benzodiazepine receptor binding
by isoflurane. Anesthesiology 2001;95:585-
93.
91. Franks NP, Dickinson R, de Sousa SL,
Hall AC, Lieb WR. How does xenon produce
anaesthesia? Nature 1998;396:324-6.
92. Daniels S, Roberts RJ. Post-synaptic
inhibitory mechanisms of anaesthesia: gly-
cine receptors. Toxicol Lett 1998;100-101:
71-6.
93. Cheng G, Kendig JJ. Enflurane directly
depresses glutamate AMPA and NMDA cur-
rents in mouse spinal cord motor neurons
independent of actions on GABAA or glycine
receptors. Anesthesiology 2000;93:1075-84.
94. Flood P, Ramirez-Latorre J, Role L. Alpha
www.nejm.org may 22, 2003
4 beta 2 neuronal nicotinic acetylcholine
receptors in the central nervous system are
inhibited by isoflurane and propofol, but
alpha 7-type nicotinic acetylcholine recep-
tors are unaffected. Anesthesiology 1997;86:
859-65.
95. Raines DE, Claycomb RJ. Nonhaloge-
nated anesthetic alkane and perhaloge-
nated nonimmobilizing alkanes inhibit
alpha(4)-beta(2) neuronal nicotinic acetyl-
choline receptors. Anesth Analg 2002;95:
573-7. [Erratum, Anesth Analg 2002;95:
869.]
96. Cordero-Erausquin M, Marubio LM,
Klink R, Changeux JP. Nicotinic receptor
function: new perspectives from knockout
mice. Trends Pharmacol Sci 2000;21:211-7.
97. Novalija E, Hogan QH, Kulier AH, Tur-
ner LH, Bosnjak ZJ. Effects of desflurane,
sevoflurane and halothane on postinfarc-
tion spontaneous dysrhythmias in dogs.
Acta Anaesthesiol Scand 1998;42:353-7.
98. Stadnicka A, Bosnjak ZJ, Kampine JP,
Kwok WM. Modulation of cardiac inward
rectifier K(+) current by halothane and iso-
flurane. Anesth Analg 2000;90:824-33.
99. Camara AK, Begic Z, Kwok WM, Bosn-
jak ZJ. Differential modulation of the cardiac
L- and T-type calcium channel currents by
isoflurane. Anesthesiology 2001;95:515-24.
100. Dodson BA, Braswell LM, Miller KW.
Barbiturates bind to an allosteric regulatory
site on nicotinic acetylcholine receptor-rich
membranes. Mol Pharmacol 1987;32:119-
26.
101. Wood SC, Tonner PH, de Armendi AJ,
Bugge B, Miller KW. Channel inhibition by
alkanols occurs at a binding site on the nico-
tinic acetylcholine receptor. Mol Pharmacol
1995;47:121-30.
102. Forman SA, Miller KW, Yellen G. A dis-
crete site for general anesthetics on a postsyn-
aptic receptor. Mol Pharmacol 1995;48:574-
81.
103. Pratt MB, Husain SS, Miller KW, Cohen
JB. Identification of sites of incorporation in
the nicotinic acetylcholine receptor of a pho-
toactivatible general anesthetic. J Biol Chem
2000;275:29441-51.
104. Mihic SJ, Ye Q, Wick MJ, et al. Sites of
alcohol and volatile anaesthetic action on
GABA(A) and glycine receptors. Nature 1997;
389:385-9.
105. Jenkins A, Greenblatt EP, Faulkner HJ,
et al. Evidence for a common binding cavity
for three general anesthetics within the
GABAA receptor. J Neurosci 2001;21:RC136.
106. Pocock G, Richards CD. Cellular mech-
anisms in general anesthesia. Br J Anaesth
1991;66:116-28.
107. MacIver MB, Kendig JJ. Anesthetic
effects on resting membrane potential are
voltage-dependent and agent-specific.
Anesthesiology 1991;74:83-8.
108. Patel AJ, Honore E. Anesthetic-sensi-
tive 2P domain K+ channels. Anesthesiology
2001;95:1013-21.
109. Vahle-Hinz C, Detsch O. What can in
2123
 drug therapy

vivo electrophysiology in animal models tell
us about mechanisms of anaesthesia? Br J
Anaesth 2002;89:123-42.
110. Antkowiak B. In vitro networks: corti-
cal mechanisms of anaesthetic action. Br J
Anaesth 2002;89:102-11.
111. Jefferys JG, Traub RD, Whittington MA.
Neuronal networks for induced ‘40 Hz’
rhythms. Trends Neurosci 1996;19:202-8.
112. Stucke AG, Stuth EA, Tonkovic-Capin V,
et al. Effects of sevoflurane on excitatory neu-
rotransmission to medullary expiratory neu-
rons and on phrenic nerve activity in a decer-
ebrate dog model. Anesthesiology 2001;95:
485-91.
113. Stuth EA, Krolo M, Stucke AG, et al.
Effects of halothane on excitatory neu-
rotransmission to medullary expiratory neu-
rons in a decerebrate dog model. Anesthesi-
ology 2000;93:1474-81.
114. Brown K, Aun C, Stocks J, Jackson E,
Mackersie A, Hatch D. A comparison of the
respiratory effects of sevoflurane and halo-
thane in infants and young children. Anes-
thesiology 1998;89:86-92.
115. Nash HA. In vivo genetics of anaes-
thetic action. Br J Anaesth 2002;89:143-55.
116. Campbell DB, Nash HA. Use of Dro-
sophila mutants to distinguish among vola-
tile general anesthetics. Proc Natl Acad Sci
U S A 1994;91:2135-9.
117. Gamo S, Dodo K, Matakatsu H, Tanaka
Y. Molecular genetical analysis of Drosoph-
ila ether sensitive mutants. Toxicol Lett 1998;
100-101:329-37.
118. Hartman PS, Ishii N, Kayser EB, Mor-
gan PG, Sedensky MM. Mitochondrial muta-
tions differentially affect aging, mutability
and anesthetic sensitivity in Caenorhabditis
elegans. Mech Ageing Dev 2001;122:1187-
201.
119. Kayser B, Rajaram S, Thomas S, Mor-
gan PG, Sedensky MM. Control of anes-
thetic response in C. elegans. Toxicol Lett
1998;100-101:339-46.
120. van Swinderen B, Saifee O, Shebester
L, Roberson R, Nonet ML, Crowder CM. A
neomorphic syntaxin mutation blocks vola-
tile-anesthetic action in Caenorhabditis ele-
gans. Proc Natl Acad Sci U S A 1999;96:
2479-84.
121. Koblin DD, Dong DE, Deady JE, Eger
EI II. Selective breeding alters murine resist-
ance to nitrous oxide without alteration in
synaptic membrane lipid composition. Anes-
thesiology 1980;52:401-7.
122. Koblin DD, Deady JE. Anaesthetic
requirement in mice selectively bred for dif-
ferences in ethanol sensitivity. Br J Anaesth
1981;53:5-10.
123. Sonner JM, Gong D, Eger EI II. Natu-
rally occurring variability in anesthetic po-
tency among inbred mouse strains. Anesth
Analg 2000;91:720-6.
124. Washburn MS, Numberger M, Zhang
S, Dingledine R. Differential dependence on
GluR2 expression of three characteristic fea-
tures of AMPA receptors. J Neurosci 1997;
17:9393-406.
125. Joo DT, Gong D, Sonner JM, et al.
Blockade of AMPA receptors and volatile
anesthetics: reduced anesthetic requirements
in GluR2 null mutant mice for loss of the
righting reflex and antinociception but not
minimum alveolar concentration. Anesthe-
siology 2001;94:478-88.
126. Carlson BX, Hales TG, Olsen RW.
GABAA receptors and anesthesia. In: Yaksh
TL, Lynch C III, Zapol WM, Maze M, Bie-
buyck JF, Saidman LJ, eds. Anesthesia: bio-
logic foundations. Philadelphia: Lippincott–
Raven, 1997:259-75.
127. Wong SM, Cheng G, Homanics GE,
Kendig JJ. Enflurane actions on spinal cords
from mice that lack the beta3 subunit of the
GABAA receptor. Anesthesiology 2001;95:
154-64.
128. Homanics GE, DeLorey TM, Firestone
LL, et al. Mice devoid of gamma-aminobu-
tyrate type A receptor beta3 subunit have epi-
lepsy, cleft palate, and hypersensitive behav-
ior. Proc Natl Acad Sci U S A 1997;94:4143-8.
129. Quinlan JJ, Homanics GE, Firestone
LL. Anesthesia sensitivity in mice that lack
the beta3 subunit of the gamma-aminobu-
tyric acid type A receptor. Anesthesiology
1998;88:775-80.
130. Homanics GE, Ferguson C, Quinlan JJ,
et al. Gene knockout on the alpha6 subunit
of the gamma-aminobutyric acid type A
receptor: lack of effect on responses to etha-
nol, pentobarbital, and general anesthetics.
Mol Pharmacol 1997;51:588-96.
131. Korpi ER, Mihalek RM, Sinkkonen ST,
et al. Altered receptor subtypes in the fore-
brain of GABA(A) receptor delta subunit-
deficient mice: recruitment of gamma 2
subunits. Neuroscience 2002;109:733-43.
132. Mohler H, Fritschy JM, Rudolph U.
A new benzodiazepine pharmacology. J Phar-
macol Exp Ther 2002;300:2-8.
133. Jurd R, Arras M, Lambert S, et al. Gen-
eral anesthetic actions in vivo strongly atten-
uated by a point mutation in the GABAA
receptor beta3 subunit. FASEB J 2003;17:
250-2.
Copyright © 2003 Massachusetts Medical Society.

personal archives in the journal online

Individual subscribers can store articles and searches using a new feature on the
Journal’s Web site (www.nejm.org) called “Personal Archive.” Each article and search
result links to this feature. Users can create personal folders and move articles into
them for convenient retrieval later.

2124 n engl j med 348;21 www.nejm.org may 22, 2003

Downloaded from www.nejm.org on August 13, 2005 . This article is being provided free of charge for use in Colombia.
Copyright © 2003 Massachusetts Medical Society. All rights reserved.