Secondary Logo

Journal Logo

Liaisons dangereuses? General anaesthetics and long-term toxicity in the CNS

Perouansky, M.*

European Journal of Anaesthesiology: February 2007 - Volume 24 - Issue 2 - p 107–115
doi: 10.1017/S0265021506001165

We do not know how general anaesthetics cause their desired effects. Contrary to what has been thought until relatively recently, the clinical state of anaesthesia consists of multiple components that are mediated via interaction of the anaesthetic drugs with different targets on the molecular, the cellular, the network and the structural–anatomical levels. The mechanisms by which some of these drugs induce the different components of ‘anaesthesia’ may be rather specific: discrete mutations of single amino acids in specific proteins profoundly affect the ability of certain anaesthetics to achieve specific endpoints. Despite this potential specificity, inhalational anaesthetics are present in the body at very high concentrations during surgical anaesthesia. Due to their lipid solubility, general anaesthetics dissolve in every membrane, penetrate into organelles and interact with numerous cellular structures in multiple ways. A priori, it is therefore not unreasonable to assume that these drugs have the potential to cause insidious changes in the body other than those acute and readily apparent ones that we routinely monitor for. Some changes may wane within a short time after removal of the drug (e.g. the suppression of immune cell function). Others may persist after complete removal of the drug and even become self-propagating (e.g. spread of malignant cells, the β-oligomerization of proteins), still others may be irreversible (e.g. the induction of apoptosis in the central nervous system) but of unclear significance. This article will focus on evidence for anaesthetic toxicity in the central nervous system, which appears to be susceptible to anaesthetic neurotoxicity primarily at the extremes of ages but via different pathways: in the neonate, during the period of most intense synaptogenesis, anaesthetics can induce excessive apoptosis; in the aging brain subtle cognitive dysfunction can persist long after clearance of the drug and processes resembling neurodegenerative disorders may be accelerated. At all ages, anaesthetics affect gene expression regulating protein synthesis in poorly understood ways. While it seems reasonable to assume that the vast majority of our patients completely restore homeostasis after general anaesthesia, exposure to these drugs probably has more profound and longer-lasting effects on the brain than heretofore imagined.

*University of Wisconsin, Department of Anesthesiology, Madison, WI, USA

Correspondence to: Misha Perouansky, Department of Anesthesiology, University of Wisconsin, Madison, WI, USA. E-mail:

Accepted for publication 16 June 2006

First published online 8 August 2006

Back to Top | Article Outline


Genome: the whole hereditary information of an organism that is encoded in the DNA (or RNA for some viruses).

Genomics: the study of the global properties of genomes (while genetics studies the properties of single genes or groups of genes).

Proteome: the complement of proteins expressed by a genome at a certain point in time.

Proteomics: the quantitative and qualitative analysis of the proteome over time and under varying conditions.

Glutamate: principal excitatory neurotransmitter in the mammalian CNS, activates ionotropic and metabotropic receptors. The three classes of ionotropic channels are named after their selective exogenous agonists (AMPA, kainate and NMDA).

GABA: γ-aminobutyric acid, the principal inhibitory neurotransmitter, like glutamate it activates ionotropic (GABAA and GABAC ) as well as metabotropic (GABAB receptors). During early postnatal development in rodents (when functional AMPA receptors are lacking), GABAA receptors are excitatory and facilitate activation of NMDA receptors [1].

Back to Top | Article Outline


Anaesthesiologists are used to thinking on a short time scale: the drugs we use are some of the most potent and fastest acting that are available in clinical medicine. Profound changes in fundamental physiological parameters take place quickly and decisions with potentially far-reaching consequences have to be frequently made on very brief notice. Conversely, we are also used to thinking that the effects of our pharmacological interventions dissipate almost as quickly as they arise and without long-term sequelae. Nowhere is the change of state as profound as in the organ that is most closely associated with our identity – the brain: deep coma, accompanied by suppression of reflexes exceeding that required for the diagnosis of brain death, gives way to normal consciousness. Unresponsiveness is followed by alertness. Complete amnesia is gradually replaced by normal memory function.

The belief that general anaesthetics are truly miraculous drugs with profound, instantaneous but fully reversible effects on the most critical organs is shared by the majority of healthcare professionals. Implicit in this belief is also the notion that once the drugs have been cleared, the body returns to the same state that it occupied before anaesthesia.

Recent laboratory data indicate that it is time for a cautious reassessment of this assumption. It appears that the drugs that brought about a revolution in medical practice more than 150 years ago may, in addition to their profound immediate effects, also have subtle but measurable long-term consequences. The purpose of this review is to provide the clinician with an overview of the current research into effects of anaesthetics on the brain that appear to have consequences that persist beyond the complete removal of the drugs.

Back to Top | Article Outline

The altered brain

Anaesthetics and brain genomics

More than a decade ago, the capacity of general anaesthetics to induce changes in gene expression in the brain was noted. Initially, only changes in the highly reactive acute early genes c-fos and c-jun were reported [2]. These acute, anaesthetic-induced changes were primarily considered to be a confounding problem for investigators and recommendations for the use of the optimal anaesthesia protocol for experimental purposes were based on the findings of differential anaesthetic drug effects on the expression of these genes [3]. Since then, effects on gene expression during and/or immediately following anaesthesia have been observed with newer anaesthetic agents, in other vital organs and with genes other than the immediate early genes [4]. Recent preliminary results indicate that changes in gene expression may also persist long beyond the time frame required to clear the anaesthetic drug. In the hippocampus of aged rats the expression of numerous genes was affected 48 h after an inhalational anaesthetic [5]. The significance of the majority of observed changes on the gene expression level remains, however, unclear. Changes in RNA expression are common and do not necessarily imply quantitative or qualitative changes in protein synthesis. Therefore changes in gene expression do not in themselves, prove that the organism underwent ‘constitutive’ changes and do not answer the question whether or not the organism that emerges from general anaesthesia without any apparent insult to its homeostasis is ‘identical’ to the organism that was anaesthetized.

Back to Top | Article Outline

Anaesthetics and brain proteomics

This question has been recently addressed for the first time using the proteomic approach. The purpose of such an analysis is to gain insight into the functional state of an organism at a certain point in time. It is a task of staggering complexity: the human genome contains 30–40 thousand genes. The number of proteins, however, is thought to exceed 1 million. Many factors account for the difference between gene number and protein diversity. Two important ones are alternate splicing of the transcriptional unit and post-translational modifications of the assembling protein. Thus, multiple proteins can be synthesized from a single gene template. However, not all proteins are present in detectable amounts at every single point in time. The proteome therefore resembles a snapshot out of the constantly changing movie of life. Detailed analysis of large numbers of proteins (including isoforms and modifications) was made possible by the advent of high-throughput methodologies such as two-dimensional polyacrylamide gel electrophoresis and mass spectrometry (for review [6]). These methods have now been used to analyse the effect of anaesthesia on the mammalian proteome. Fuetterer and colleagues [7] exposed rats to 3 h of 5.7% desflurane in air and prepared whole-brain homogenates immediately after the exposure as well as 24 and 72 h thereafter. The cytosolic proteins obtained at these time points were compared with proteins obtained from rats that were not anaesthetized (except very briefly for sacrifice). Out of the thousands of proteins that are present simultaneously in a living rat brain, 263 ‘spots’ representing individual proteins met the inclusion criteria for further analysis. Even within this limited sample, the researchers found that the abundance of a number of proteins was changed (either decreased or increased) in the brain of the rats exposed to desflurane for up to 72 h after exposure. The change ranged from a decrease to 60% to an increase to 179% of control. Although no firm conclusions about the long-term consequences or functional implications of these changes can be drawn from this work – single doses for single durations of single agents do not support firm or generalizable deductions – they do warrant attention as they indicate that ‘constitutive’ changes are indeed induced by anaesthetics. In coming years, it will be possible to extend these investigations using alternatives to gel-based detection methods with increased sensitivity and reduced variation, and this may allow the analysis of the functional consequences of the reported fluctuations in protein expression. Nevertheless, the results of these experiments indicate that on the molecular level, general anaesthesia is not as reversible a condition as the outward appearance suggests. Under certain conditions, these lasting changes may be used to the advantage of the anaesthetized subject: protection of the myocardium from ischemia by anaesthetic preconditioning is associated with changes in the myocardial proteome [8] and has been an unexpected benefit derived from a class of drugs used to achieve completely unrelated endpoints. Recent data suggest that the benefits of anaesthetic preconditioning may also extend to nervous tissue [9]. However, these discoveries should remind us that familiar drugs can have unexpected side-effects and that not all of them are bound to be beneficial.

Back to Top | Article Outline

Anaesthetic neurotoxicity –in vitro

Mechanisms of neurodegeneration

One proposed mechanism of Alzheimer's disease (AD) (and possibly of other neurodegenerative disorders) that is supported by extensive experimental data is that uncontrolled oligomerization (microaggregation) of peptides or proteins that are normally present in the brain leads to neurotoxic effects. The suspected peptide in AD is the amyloid β peptide (Aβ), 35–42 amino acids in length (e.g. Aβ42), that is constitutively released after proteolytic cleavage of the amyloid precursor protein (APP). Amorphous deposits of Aβ, i.e. not the fibrillar form, can be found in large amounts in the brains of AD patients but also, to a lesser extent, in normal-aged humans and their toxic potential is not known. Under certain environmental circumstances, however, these peptides are capable of changing their secondary structure from α-helix to β-sheet and self-assemble to oligomers of varying and increasing sizes. The oligomers produce the characteristic fibrils, the building blocks of the pathognomonic AD plaques. The other typical AD lesion, the neurofibrillary tangle, is composed of bundles of the microtubule-associated protein τ. While it is not entirely clear how and what products of oligomerization cause neurodegeneration, multiple lines of evidence bolster the ‘amyloid (or more precisely Aβ) hypothesis of AD’ (reviewed by [10,11]): various conditions that lead to increased Aβ formation increased the likelihood and the severity of damage to the central nervous system (CNS); Aβ was neurotoxic, inhibited the induction of long-term potentiation (LTP, a cellular form of learning and memory), while washout of Aβ protein or prevention of oligomerization of Aβ rescued LTP and neutralization of Aβ prevented certain manifestations of AD in some models [12]. Current thinking within the framework of the Aβ hypothesis envisages the neurotoxic element to be the intermediate-sized oligomer and not the mature fibril. Therefore, the possibility that the physicochemical nature of inhalational anaesthetics could favor oligomerization of Aβ as they have been shown to do for other proteins is intriguing.

Back to Top | Article Outline

Anaesthetics – molecular catalysts of neurodegeneration?

The interaction of anaesthetic molecules and Aβ peptides was tested recently by Eckenhoff and colleagues in a cell-free experimental system [13]. The presence of halothane and of isoflurane in the protein suspension significantly accelerated the oligomerization of Aβ42 and led to an increase in the total amount of oligomerized peptide (note, however, that the lowest concentration tested was 1 mmol, that is more than three times minimal alveolar concentration (MAC)). This effect was specific for the substrate (human serum albumin, a hydrophobic protein used as control, was not affected) and for inhalational drugs: propofol and alcohol, when tested at ‘clinical’ concentrations, mildly inhibited oligomerization (but enhanced oligomerization at very high concentrations). Removal of halothane from the protein suspension after 1 h of incubation did not reverse this process immediately: enhanced amounts of oligomers were detectable for up to 3 days thereafter. In a second series of experiments, the authors examined the toxicity of the oligomer– anaesthetic combination in rat pheochromocytoma cells, a cell type that, being of neural crest origin, is frequently used as a model for nerve cells. Toxicity was assayed by measuring the release of lactate dehydrogenase (LDH) from cultured pheochromocytoma cells incubated with 15 μmol Aβ42 in the presence of different anaesthetic drugs for 72 h. Neither isoflurane nor halothane, when given alone, increased LDH release, indicating lack of toxicity in this model. By contrast, incubation of pheochromocytoma cells with Aβ42 increased LDH release above control levels. Addition of the equivalent of 1 MAC of the volatile agents to Aβ42 increased its toxicity while addition of propofol or ethanol did not. The researchers concluded that under their experimental conditions volatile anaesthetics enhanced the formation of amyloid and its toxicity in cell culture. If similar processes occurred in vivo, volatile anaesthetics could lead to long-lasting increases of Aβ42, a neurotoxic form of amyloid in the brain of susceptible subjects. Recently, this same group presented preliminary data indicating that the AD-promoting PS-1 mutation, transfected into pheochromocytoma cells, rendered these cells more susceptible to isoflurane-induced toxicity [14].

Back to Top | Article Outline

Anaesthetic neurotoxicity –in vivo

The most extensive clinical trials published to date [15–17] have neither proven nor excluded a direct causal link between anaesthetics and cognitive impairment. Therefore, some investigators have turned to animal models to test the hypothesis that general anaesthetics can induce changes in the brain that outlast their physical presence there and that these changes lead to long-lasting, measurable functional consequences.

Back to Top | Article Outline

The mature brain: adult vs. aged

General anaesthetics affect learning and memory. Temporary amnesia is a highly desirable component of the anaesthetic state that can be reliably achieved with general anaesthetics. Prolonged memory impairment, by contrast, is one essential component of postoperative cognitive dysfunction (POCD).

In the early 1990s, laboratory studies of anaesthetic interaction with memory reported a seemingly paradoxical improvement of memory consolidation. Adult mice anaesthetized with halothane, isoflurane or enflurane immediately after aversive conditioning had improved performance compared to unanaesthetized controls, when tested on the same task 24 hours later [18]. The mechanisms underlying the observed memory improvement by anaesthesia were not further investigated, but it was noted that in these experiments anaesthetic drugs were present in the brain at anaesthetic concentrations during a time window that might be important for memory consolidation on the cellular–molecular level. Improvement of memory by anaesthetics not explainable by the same mechanism was observed in the laboratory of Dr G. Crosby where researchers have worked to establish an animal model of long-term anaesthetic effects on cognitive performance [19]. The prototypical experiment consisted of evaluating the effects of a standardized anaesthetic (2 h of 1.2% isoflurane/70% N2O/30% oxygen) on performance in a 12-arm radial arm maze (RAM, a standard test to evaluate spatial orientation skills). The animals were either adult (6 months old) or aged (18–20 months old) Fischer 344 rats (median life expectancy 26 months) and were randomly assigned to either the test or the age-matched control group. The experimental protocols differed in the relative timing of anaesthesia (before or after training), the degree of training and the duration of the observation period after anaesthesia. In contrast to the work by Komatsu [18], when training followed anaesthesia, it was separated from anaesthesia by a time interval sufficient to eliminate the drugs. In their initial publication, Culley and colleagues reported that anaesthesia administered 24 h after the training had differential effects in adult vs. aged rats: it improved performance in the former and impaired it in the latter. These effects remained measurable for 3 weeks and were undetectable by 8 weeks after anaesthesia [19]. However, at the time of anaesthesia the adult animals (who learned faster) were overtrained compared to the old animals and learning continued to take place during the post-anaesthesia testing sessions. Therefore, this experimental protocol, did not allow for a separation between effects on acquisition of new information (learning) vs. retrieval of stored information (memory). When, by contrast, the effect of anaesthesia was tested on the ability to acquire new memories (training/testing took place only after anaesthesia), the same investigators found that learning was impaired in both adult and aged animals even when the training began for up to 2 weeks after anaesthesia, i.e. the impairment lasted for considerably longer than predicted by the pharmacokinetics of the drugs used [20,21]. Interestingly, the researchers were not able to demonstrate any age-by-anaesthesia interactions, that is both age groups were equally impaired. The authors attributed this apparent inconsistency with their earlier study to differences in the protocol: retrieval of stored information in the first study vs. acquisition of new memory (more susceptible to disruption and therefore affected in both age groups) in the later ones. In summary, this group of investigators concluded that inhalational anaesthetics had effects on spatial learning and memory that outlasted their physical presence in the body. The mechanism(s) of these protracted memory effects and whether or not they are related in any way to either in vitro or to clinical observations is unclear, and the results have not yet been replicated in other laboratories or using other memory paradigms. Nevertheless, the findings indicate that the aged rodent may prove to be a useful model for studying the mechanisms of anaesthetic-induced cognitive dysfunction. Along with the intriguing Aβ-neurodegenerative hypothesis, other mechanisms interlinking anaesthetic drugs and persistent neurological impairment should be explored, for example a possible link between the effects of anaesthetics on mitochondrial function [22,23] and CNS dysfunction [24].

Back to Top | Article Outline

The immature brain

Neurotoxicity, apoptosis and alcohol.

Among the proteins expressed in the brain whose function is affected by anaesthetic drugs at concentrations achieved during clinical anaesthesia are receptors for the most abundant excitatory and inhibitory neurotransmitters in the mammalian CNS: l-glutamate and γ-aminobutyric acid (GABA). Both transmitter systems play important roles also in the developing brain. All clinically used general anaesthetics either enhance GABAA receptors or block N-methyl-d-aspartate (NMDA) receptors or do both (it is not known with certainty, however, to what degree these effects contribute to the clinical phenomenon of general anaesthesia). In the adult and the developing brain, excessive activation of l-glutamate receptors, primarily of the NMDA type is associated with neurotoxicity. In contrast to the mature brain, however, it has been recently discovered that transient pharmacological blockade of NMDA receptors in the developing rodent brain causes excessive neuronal apoptosis. This excess is seen only in brain areas that undergo physiological apoptosis in normal development and only if NMDA receptors are blocked at specific time points within a period of accelerated synaptogenesis [25]. It appears that during this brain growth spurt, neurons that do not receive sufficient excitation are doomed to apoptotic degeneration [26].

Ethanol, in the same model, causes neuronal degeneration that is even more widespread than that caused by selective NMDA receptor blockade. While alcohol is not used clinically as an anaesthetic drug, it is important to note that alcohols are model anaesthetics that share targets and mechanisms with clinically used anaesthetics. An apoptogenic potential of alcohol on the developing human brain (synaptogenesis in human beings lasts throughout the last trimester of pregnancy and continues postnatally for at least two years) would be therefore clinically relevant. The widespread alcohol-induced apoptosis could be attributed to the fact that alcohol in addition to blocking NMDA receptors also enhances GABAA receptors, i.e. developing neurons are ‘silenced’ via two pathways and are therefore more likely to commit to apoptosis. Similarly to the NMDA blockers, alcoholic neurotoxicity was only observed when the alcohol was administered during the critical developmental period of intense synaptogenesis [27]. In the rat, synaptogenesis is most intense in the last 2 days of pregnancy (gestation lasts 21 days in the rat) and the first two postnatal weeks. Within this time window, different brain regions show peak sensitivity to alcohol's proapoptotic effects at different times. When administered on the 7th postnatal day (P7), alcohol caused a approximately 20-fold increase in the density of degenerating cells in the CA1 area of the hippocampus and increases of the same order of magnitude were found in the thalamus, the septum, basal ganglia and some cortical areas [27]. Alcoholic neurodegeneration was also associated with a decreased brain mass leading the authors to conclude that neurodegeneration mediated by alcohol's effects on NMDA and GABAA receptors could explain some aspects of the clinical presentation of fetal alcohol syndrome. It should be noted, however, that alcohol also affects other aspects of neuronal development in detrimental ways, for example interference with second messenger signalling leading to aberrant neuronal migration in the fetal brain [28].

Back to Top | Article Outline

Anaesthesia and apoptosis.

This issue has been addressed recently in a series of experiments that approximated clinical conditions. Jevtovic-Todorovic and colleagues [29] anaesthetized rat pups on P7 with combinations of midazolam, isoflurane and nitrous oxide sufficient to maintain a surgical plane of anaesthesia for 6 h. Controls were exposed to 6 h of mock anaesthesia. After recovery, the animals were divided into three randomly selected groups. The first group was used for histopathological studies. The second group was used for behavioural studies in adulthood, which involved evaluating the rats on several tests over a 160-day period. The third group was used to study hippocampal plasticity in vitro (a model for learning and memory) at P29–P33. Histopathology revealed that isoflurane (0.75–1.5%) caused a dose-dependent increase in apoptotic neurodegeneration. Midazolam (3–9 mg kg−1) and N2O (50–150% in a hyperbaric chamber) when given alone did not cause increased rates of apoptosis compared to control animals. However, when midazolam was followed by maintenance with isoflurane (double cocktail) the damage caused by the latter was increased primarily in the thalamus and the parietal cortex. The damage was further augmented by adding N2O to the maintenance phase of anaesthesia (triple cocktail: midazolam 9 mg kg−1, isoflurane 0.75%, N2O 75%). The triple cocktail caused widespread more than 15-fold increases in the number of apoptotic neurons. When the hippocampi (a structure important for memory formation) from rats treated with the triple cocktail at P7 were evaluated electrophysiologically at P29– P33, it was found that the ability of neuronal connections to express LTP was severely reduced, suggesting that learning and memory may also be impaired. This deficit was confirmed in behavioural tests: animals exposed to the triple cocktail as neonates had measurable deficits in spatial memory tests when tested in the water maze and the radial arms maze as adults. Recently, this group of researchers proposed that anaesthetics activate both the intrinsic and the extrinsic apoptotic pathways, depending on the length of the exposure [30]. In neonatal rats, the intrinsic pathway is activated within 2 h of anaesthesia exposure, while activation of the extrinsic pathway occurs later – within 6 h. It was also noted that the susceptibility to induction of apoptosis decreased dramatically from the 7th to the 14th postnatal day, coinciding with reduced rate of synaptogenesis.

Back to Top | Article Outline

Clinical relevance.

These studies and their potential extrapolations to clinical anaesthesia in children generated a vigorous debate at scientific meetings and in the anaesthesia literature. Indeed, one manuscript fully dedicated to critique and discussion of these experiments and a number of ‘point–counterpoint’ exchanges were published. They are interesting to review as they illustrate the collision of intriguing laboratory findings with long-established clinical practices. The critique and discussion by Anand and Soriano included both methodological and conceptual issues [31]. With respect to methodology it was suggested that the reported histopathological damage and neurological deficits could be attributed to factors other than the anaesthetic drugs (e.g. hypoxia or malnutrition). A recent investigation directly addressed the issue of metabolic homeostasis in a neonatal mouse model. Loepke and colleagues found that mouse pups anaesthetized with 1.8% isoflurane developed marked hypoglycaemia within 90 min (to below 40 mg% in 50% of animals), while control pups separated from the dam for the same amount of time did not [32]. Hypoglycaemia of this severity has been shown to cause neuronal damage in adult rodents in some models, although primarily of the necrotic [33] as opposed to the apoptotic type described by Jevtovic-Todorovic. Moreover, the neonate brain is more adept at using alternate fuel sources and therefore may be more resistant to hypoglycaemia-induced damage [34]. In addition to posing the interesting question of how isoflurane induces hypoglycaemia in neonates (hyperglycaemia is seen in adult rodents under isoflurane), this finding also offers an additional potential mechanism for the neurodegeneration attributed solely to anaesthetics by Jevtovic-Todorovic and colleagues [29]. A number of considerations, however, indicate that hypoglycaemia is not the principal factor inducing apoptosis: isoflurane-induced hypoglycaemia explains neither the potentiation of neuronal damage by co-administration of multiple anaesthetics nor the neurotoxic effect of ketamine or alcohol administered alone. Moreover, preliminary results indicate that neurodegeneration is also observed when guinea pig fetuses are exposed to anaesthetics in utero, when they should be far less susceptible to developing hypoglycaemia as the dam remains normoglycemic throughout the experiment [35].

Moreover, recently, isoflurane neurotoxicity was tested in cultured slices from rat hippocampi harvested at 3, 7 and 14 days postnatally. This ‘in vitro’ model has the advantage of eliminating homeostatic confounding factors present in vivo (hypoxemia, acidosis, hypoglycaemia, etc.). The results were in general agreement with the in vivo findings, even though in vivo the hippocampus is not the structure most affected by excessive induction of apoptosis. Isoflurane 1.5% for 5 h caused a significant increase in the number of apoptotic neurons in slices obtained from 7-day-old animals. Cultures obtained from either younger or older animals or exposed to isoflurane for shorter time periods were less affected [36]. A conclusion that can be drawn from this study is that hypoglycaemia is not the sole mechanism by which anaesthetics can induce neuronal apoptosis. Nevertheless, the role of hypoglycaemia in the anaesthetic-induced apoptotic neurodegeneration requires further clarification.

The conceptual critique by Anand and Soriano [31] questioned the appropriateness of drawing relevant conclusions for clinical scenarios in humans from data obtained in rodents. The pivotal issue here is the different developmental and lifetime scales between the two species. In other words, are 6 h of drug exposure during synaptogenesis such an extreme insult to the rodent brain that it precludes any extrapolation to clinical anaesthesia in humans? There is no final answer but the following data (from rodents) indicate that toxicity occurs with lesser insults as well. Detailed dose-response data for anaesthetic cocktails of predominantly GABAA-enhancing drugs are not available. The ‘triple cocktail’ (GABA-enhancing and NMDA-blocking drugs), however, appears to activate the intrinsic apoptotic pathway already within 2 h [30]. Ketamine alone (a potent NMDA receptor blocker) at sub-anaesthetic doses induced dose-dependent, incremental apoptosis and identifiable toxicity [37]. Recently the neurotoxic potential of a single ‘mildly’ anaesthetic dose (defined as preservation of pain response and righting reflex) of both ketamine and midazolam (individually and combined) in rat pups has also been demonstrated [38]. Before conclusions for clinical practice can be drawn, however, the hypothesis that anaesthetic agents cause neurodegeneration under clinical circumstances requires testing in species with different patterns of synaptogenesis. Preliminary results from guinea pig (synaptogenesis in this species takes place prenatally and over a much longer time period than in rats) [35] and piglet (synaptogenesis is prolonged and spans pre- and postnatal time periods) [39] models indicate that these animals are also susceptible to anaesthetic neurotoxicity if exposed to midazolam, isoflurane and nitrous oxide pre- (guinea pigs) and postnatally (piglets). A non-human primate model may provide particularly useful information but is also the most difficult to establish.

Back to Top | Article Outline


Laboratory evidence of anaesthetic-induced lasting changes in or potentially affecting the CNS include the following:

  • Increased β-oligomerization of proteins after exposure to and increased cytotoxicity of the oligomers in the presence of volatile anaesthetics [13].
  • Brief exposure to desflurane causes changes in protein expression in the brain that far outlast the presence of the drug [7].
  • Anaesthetics induce neuronal degeneration in the developing brain, the effect of combined application of drugs acting via different mechanisms has more than additive effects [29].
  • Anaesthetics have the potential to activate both the intrinsic and the extrinsic apoptotic pathways [30].
  • Neonatal exposure to anaesthetic drugs has measurable effects on learning in adulthood [29].
  • Exposure of aged rats to inhalational anaesthetics causes lasting impairment of spatial memory performance [40].
Back to Top | Article Outline

Is anaesthetic neurotoxicity a clinically relevant issue?

The International Study of Post-Operative Cognitive Dysfunction (ISPOCD) has convincingly demonstrated the existence of POCD in elderly patients as a clinical entity [16]. The only risk factor for the development of prolonged or irreversible POCD that could be clearly identified was advancing age. Why advancing age increases the likelihood remains unclear. Simple violations of homeostasis (hypotension and hypoxemia) might be expected to occur more frequently with increasing age, but have been excluded by the ISPOCD studies as the causative link between anaesthesia and POCD. One possibility, proposed more than a decade ago but never substantiated [41], linking age and anaesthesia with cognitive dysfunction is that exposure to (surgery and) anaesthesia accelerates pre-existing but heretofore latent neurodegenerative disorders (e.g. AD) that then become manifest sometime in the post-operative period as POCD or MCI (mild cognitive impairment, a prelude to AD). Could anaesthetic drugs affect the progression of neuro-degenerative disorders? Where in the pathogenetic chain of cognitive dysfunction could these small and relatively inert molecules play a role?

No data obtained from humans suggest the existence of anaesthetic neurotoxicity at the other extreme of age, the neonate. Nevertheless, there is laboratory evidence in rodent models pointing towards the possibility that the immature CNS might be susceptible to deleterious effects induced by commonly used anaesthetic drugs, and that clinical evidence for such an effect might simply await a concerted effort at identifying it, analogous the ISPOCD study.

In the light of the existing evidence, the question is not so much whether long-lasting, anaesthetic-induced changes in the CNS do occur but whether they have any identifiable or preventable deleterious impact. Convincing clinical evidence attesting to that is lacking but may not be obtainable due to the methodological constraints of studies in humans. Therefore, additional evidence will have to be gathered from animal models that more closely mimic clinical reality. Until proven otherwise, however, it may be prudent to assume that there are no side-effect free pharmacological interventions. The ideal of stress-free surgery may be more difficult to achieve then expected, especially in the most vulnerable patient populations.

Back to Top | Article Outline


I am grateful to Drs I. Matot and R. A. Pearce for critical reading of the manuscript.

Back to Top | Article Outline


1. Ben-Ari Y, Khazipov R, Leinekugel X, Caillard O, Gaiarsa JL. GABAA, NMDA and AMPA receptors: a developmentally regulated ‘menage a trois’. Trend Neurosci 1997; 20(11): 523–529.
2. Marota JJ, Crosby G, Uhl GR. Selective effects of pentobarbital and halothane on c-fos and jun-B gene expression in rat brain. Anesthesiology 1992; 77(2): 365–371.
3. Takayama K, Suzuki T, Miura M. The comparison of effects of various anesthetics on expression of fos protein in the rat-brain. Neurosci Lett 1994; 176(1): 59–62.
4. Hamaya Y, Takeda T, Dohi S, Nakashima S, Nozawa Y. The effects of pentobarbital, isoflurane, and propofol on immediate-early gene expression in the vital organs of the rat. Anesth Analg 2000; 90(5): 1177–1183.
5. Culley DJ, Yukhananov RY, Crosby G. Hippocampal gene expression is altered 48 hours after general anesthesia in aged rats. Anesthesiology 2004; 101: A64.
6. Atkins JH, Johansson JS. Technologies to shape the future: proteomics applications in anesthesiology and critical care medicine. Anesth Analg 2006; 102(4): 1207–1216.
7. Futterer CD, Maurer MH, Schmitt A, Feldmann RE, Kuschinsky W, Waschke KF. Alterations in rat brain proteins after desflurane anesthesia. Anesthesiology 2004; 100(2): 302–308.
8. Kalenka A, Maurer MH, Feldmann RE, Kuschinsky W, Waschke KF. Volatile anesthetics evoke prolonged changes in the proteome of the left ventricule myocardium: defining a molecular basis of cardioprotection? Acta Anaesthesiol Scand 2006; 50(4): 414–427.
9. Xiong L, Zheng Y, Wu M, Hou L, Zhu Z, Zhang X et al. Preconditioning with isoflurane produces dose-dependent neuroprotection via activation of adenosine triphosphate-regulated potassium channels after focal cerebral ischemia in rats. Anesth Analg 2003; 96(1): 233–237.
10. Hardy J, Selkoe DJ. The amyloid hypothesis of Alzheimer's disease: progress and problems on the road to therapeutics. Science 2002; 297(5580): 353–356.
11. Walsh DM, Selkoe DJ. Deciphering the molecular basis of memory failure in Alzheimer's disease. Neuron 2004; 44(1): 181–193.
12. Walsh DM, Klyubin I, Fadeeva JV, Cullen WK, Anwyl R, Wolfe MS et al. Naturally secreted oligomers of amyloid beta protein potently inhibit hippocampal long-term potentiation in vivo. Nature 2002; 416(6880): 535–539.
13. Eckenhoff RG, Johansson JS, Wei HF, Carnini A, Kang BB, Wei WL et al. Inhaled anesthetic enhancement of amyloid-beta oligomerization and cytotoxicity. Anesthesiology 2004; 101(3): 703–709.
14. Wei H, Liang G, Wei W, Li Y, Eckenhoff RG. Volatile anesthetics induce neurotoxicity differently in PC12 cells featured with Alzheimer's Disease. Anesthesiology 2005; 103: A121.
15. Mikkelsen H, Rentowl P, Rasmussen LS, Hanning CD, Moller JT. Cognitive dysfunction may persist 1–2 years after major non-cardiac surgery. Anesthesiology 1998; 89(3A): U963.
16. Moller JT, Cluitmans P, Rasmussen LS, Houx P, Rasmussen H, Canet J et al. Long-term postoperative cognitive dysfunction in the elderly: ISPOCD1 study. Lancet 1998; 351(9106): 857–861.
17. Rasmussen LS, Johnson T, Kuipers HM, Kristensen D, Siersma VD, Vila P et al. Does anaesthesia cause postoperative cognitive dysfunction? A randomised study of regional versus general anaesthesia in 438 elderly patients. Acta Anaesth Scand 2003; 47(3): 260–266.
18. Komatsu H, Nogaya J, Anabuki D, Yokono S, Kinoshita H, Shirakawa Y et al. Memory facilitation by posttraining exposure to halothane, enflurane, and isoflurane in ddN mice. Anesth Analg 1993; 76(3): 609–612.
19. Culley DJ, Baxter M, Yukhananov R, Crosby G. The memory effects of general anesthesia persist for weeks in young and aged rats. Anesth Analg 2003; 96(4): 1004–1009.
20. Crosby C, Culley DJ, Baxter MG, Yukhananov R, Crosby G. Spatial memory performance 2 weeks after general anesthesia in adult rats. Anesth Analg 2005; 101(5): 1389–1392.
21. Culley DJ, Baxter MG, Crosby CA, Yukhananov R, Crosby G. Impaired acquisition of spatial memory 2 weeks after isoflurane and isoflurane–nitrous oxide anesthesia in aged rats. Anesth Analg 2004; 99(5): 1393–1397.
22. Miro O, Barrientos A, Alonso JR, Casademont J, Jarreta D, Urbano-Marquez A et al. Effects of general anaesthetic procedures on mitochondrial function of human skeletal muscle. Eur J Clin Pharmacol 1999; 55(1): 35–41.
23. Kayser EB, Morgan PG, Sedensky MM. GAS-1: a mitochondrial protein controls sensitivity to volatile anesthetics in the nematode Caenorhabditis elegans. Anesthesiology 1999; 90(2): 545–554.
24. Morgan PG, Hoppel CL, Sedensky MM. Mitochondrial defects and anesthetic sensitivity. Anesthesiology 2002; 96(5): 1268–1270.
25. Ikonomidou C, Bosch F, Miksa M, Bittigau P, Vockler J, Dikranian K et al. Blockade of NMDA receptors and apoptotic neurodegeneration in the developing brain. Science 1999; 283(5398): 70–74.
26. Mennerick S, Zorumski CF. Neural activity and survival in the developing nervous system. Mol Neurobiol 2000; 22(1–3): 41–54.
27. Ikonomidou C, Bittigau P, Ishimaru MJ, Wozniak DF, Koch C, Genz K et al. Ethanol-induced apoptotic neurodegeneration and fetal alcohol syndrome. Science 2000; 287(5455): 1056–1060.
28. Kumada T, Lakshmana MK, Komuro H. Reversal of neuronal migration in a mouse model of fetal alcohol syndrome by controlling second-messenger signalings. J Neurosci 2006; 26(3): 742–756.
29. Jevtovic-Todorovic V, Hartman RE, Izumi Y, Benshoff ND, Dikranian K, Zorumski CF et al. Early exposure to common anesthetic agents causes widespread neurodegeneration in the developing rat brain and persistent learning deficits. J Neurosci 2003; 23(3): 876–882.
30. Yon JH, Daniel-Johnson J, Carter LB, Jevtovic-Todorovic V. Anesthesia induces neuronal cell death in the developing rat brain via the intrinsic and extrinsic apoptotic pathways. Neuroscience 2005; 135(3): 815–827.
31. Anand KJ, Soriano SG. Anesthetic agents and the immature brain: are these toxic or therapeutic? Anesthesiology 2004; 101(2): 527–530.
32. Loepke AW, McCann JC, Kurth D, McAuliffe JJ. The physiologic effects of isoflurane anesthesia in neonatal mice. Anesth Analg 2006; 102: 75–80.
33. Auer RN. Hypoglycemic brain damage. Metab Brain Dis 2004; 19(3–4): 169–175.
34. Belik J, Wagerle LC, Stanley CA, Sacks LM, Herbert DW, Delivoriapapadopoulos M. Cerebral metabolic response and mitochondrial activity following insulin-induced hypoglycemia in newborn lambs. Biol Neonate 1989; 55(4–5): 281–289.
35. Rizzi S, Yon JH, Carter LB, Jevtovic-Todorovic V. Short exposure to general anesthesia caused widespread neuronal suicide in the developing guinea pig brain. Soc Neurosci Program 2005; 241: 6.
36. Wise-Faberowski L, Zhang H, Ing R, Pearlstein RD, Warner DS. Isoflurane-induced neuronal degeneration: an evaluation in organotypic hippocampal slice cultures. Anesth Analg 2005; 101(3): 651–657 (Table).
37. Olney JW, Young C, Wozniak DF, Ikonomidou C, Jevtovic-Todorovic V. Anesthesia-induced developmental neuroapoptosis. Does it happen in humans? Anesthesiology 2004; 101(2): 273–275.
38. Young C, Jevtovic-Todorovic V, Qin YQ, Tenkova T, Wang H, Labruyere J et al. Potential of ketamine and midazolam, individually or in combination, to induce apoptotic neurodegeneration in the infant mouse brain. Brit J Pharmacol 2005; 146(2): 189–197.
39. Rizzi S, Carter LB, Jevtovic-Todorovic V. Clinically used general anesthetics induce neuroapoptosis in the developing piglet brain. Soc Neurosci Prog 2005; 251: 7.
40. Culley DJ, Baxter MG, Yukhananov R, Crosby G. Long-term impairment of acquisition of a spatial memory task following isoflurane–nitrous oxide anesthesia in rats. Anesthesiology 2004; 100(2): 309–314.
41. Houx PJ, Vreeling FW, Jolles J. Age-associated cognitive decline is related to biological life events. In: Iqbal K, McLachlan DRC, Winblad B, Wisniewski HS, eds. Alzheimer's Disease: Basic Mechanisms, Diagnosis, and Therapeutic Strategies. Chichester, UK: John Wiley & Sons, 1991: 353–358.


© 2007 European Society of Anaesthesiology