Exposure of neonates and young children to anaesthetics has long been approached with extreme caution, due to the depressant effects of these drugs on immature organ systems. Consequently, for many years, it was common practice to perform anaesthesia in these patients using only nitrous oxide and curare and withholding volatile and intravenous (i.v.) anaesthetic drugs . During the past 25 yr, major achievements have been made in our understanding of neonatal physiology and it is now well established that withholding anaesthetics during the perioperative period increase the incidence of intraoperative and postoperative complications [2–4]. Also, long-term consequences of repetitive and prolonged pain in the neonatal period will include important changes in pain sensitivity  as well as a variety of neurodevelopmental, behavioural and cognitive deficits in later childhood [6,7].
While the necessity of providing adequate anaesthesia and analgesia levels for young children undergoing surgery or other painful procedures is clearly beyond question, a growing number of laboratory observations indicate that exposure to anaesthetics might potentially have adverse effects on central nervous system (CNS) development. Indeed, currently used anaesthetics act mainly by modulating the activity of GABAergic and glutaminergic neurotransmitter systems, and it is now well established that, in addition to their role in synaptic transmission in the CNS, gamma-aminobutyric acid (GABA) and glutamate also act during development, as epigenetic factors to control important biological processes including progenitor cell proliferation, neuroblast migration and dendritic maturation [8,9]. These effects appear to be mediated through a paracrine, diffuse mode of action that contrasts with the more focused, rapid mode of operation at the synapse. The objectives of the present review is to provide a brief synopsis of how GABAergic and glutaminergic signalling contribute to shape brain development as well as to summarize accumulating evidence suggesting that anaesthetics might indeed interfere with neuronal differentiation and survival.
GABA and glutamate as early developmental factors
Neurotransmitters are primarily considered as the effectors of synaptic transmission. It is, however, important to note that they are also present in the chemical microenvironment of neural cells from the very early stages of CNS development [10,11]. In human embryos, GABA positive cells can be detected as early as the sixth gestational week , a period coinciding with the beginning of cerebral wall formation. Glutamate, the major excitatory amino acid in the adult CNS, is also one of the most abundant neurotransmitters during CNS development . In turn, neural stem cells, migrating neuroblasts and immature neurons express specific receptors for neurotransmitters [11,13]. Thus, in addition to their role in neurotransmission in mature neuronal circuits, GABA and glutamate can also serve as chemical signals to orchestrate a variety of morphogenetic events during brain development [14,15]. These effects appear to be, at least partially, mediated through a paracrine, diffuse mode of action where GABA and glutamate are non-synaptically released into the early differentiating cortical environment and activate their cogent ionotropic receptors . This type of early neurotransmitter signalling might possibly mediate a wide range of developmental effects, including proliferation, differentiation and synapse formation, and is in contrast to the more focused, rapid mode of operation at the synapse.
Proliferation of neural precursor cells is one of the first important steps during cerebral wall formation. Two distinct proliferative populations contribute to the development of the cerebral neocortex, the ventricular zone and the subventricular zone . The ventricular zone mainly generates neuronal precursors during the embryonic period and is then replaced by ependymal cells with limited proliferative capacity in adulthood . In contrast, the subventricular zone persists as a proliferative population during the whole life span and predominantly generates glial cells and a limited number and types of neurons [18,19].
The role of GABA and glutamate in progenitor cell proliferation has been extensively studied in laboratory animals at comparable developmental stages [12,20–24]. Application of GABA as well as glutamate increases cell proliferation by shortening the cell cycle in the ventricular zone, but, on the other hand, these neurotransmitters also decrease the proliferation rate of subventricular zone progenitors . This differential modulation of cell proliferation could regulate the relative contribution of ventricular and subventricular zone progenitors to neocortical growth and thus could be of utmost importance for proper CNS development [24,25].
Recent results further extend our understanding about the regulatory role of GABA on cell proliferation during development, by demonstrating the existence of non-synaptic GABAergic signalling between neuroblast and glial progenitors in the postnatal subventricular zone . According to these data, spontaneous depolarization-induced non-synaptic GABA release in neuroblasts activates GABAA receptors on adjacent astrocytes leading to a decreased proliferation of these latter cell types. The role of GABA and glutamate has also been established in the migration of newly generated neural precursor cells toward their destinations. GABA was shown to have chemotactic effects during brain maturation by guiding the migration of newly generated embryonic neurons from the ventricular zone to the cortical plate [22,23]. Glutamate, acting through the NMDA receptor, also stimulates embryonic cortical neuronal migration [22,27].
The synergistic trophic actions of GABA and glutamate during CNS development are mainly explained by the fact that, due to the high intracellular Cl− concentrations of immature neurons, activation of GABAA receptors leads to the depolarization of these cells. Thus, not only glutamate but also GABA acts as an excitatory neurotransmitter during brain development. The functional switch toward the hyperpolarizing actions of this neurotransmitter is linked to the developmental expression of the K+–Cl− cotransporter (KCC2), actively extruding intracellular Cl− from neurons . KCC2 appears in the early postnatal period in rodents, while no data about the expression pattern of this protein exist in human beings yet.
Role of GABA and glutamate in neuronal differentiation and network formation
The most active phase of neuronal differentiation, synaptogenesis and functional network formation in the rodent brain takes place between the first and third postnatal weeks (corresponding to a period extending from the last trimester of pregnancy up to the first few years of life in human beings), which closely parallels the onset of sensory input to the cortex [29,30]. Indeed, substantial evidence supports the view that afferent synaptic and network activity plays a fundamental role in shaping neuronal arbor development [31–33]. Dendrites represent the primary sites of synaptic contacts in developing neurons, and the elaboration of a highly complex and organized dendritic arbor is a prerequisite for the establishment of neuronal circuitry . While the particular morphology of CNS neurons appears to be specified genetically , these intrinsic programmes act in concert with extracellular signals during the differentiation of the dendritic arbor .
During the early stages of neuronal arbor elaboration, both GABA and glutamate act as non-synaptic trophic factors to promote differentiation. Exposure to GABA leads to increases in dendritic length, branching and in the density of synapses in several in vivo and in vitro model systems, while antagonism of the GABAA receptor, using the selective GABAA receptor antagonist bicucullin, has opposite effects . Pharmacological blockade of the NMDA-type glutamate receptors also markedly reduces dendritic growth rates .
Synaptically released GABA and glutamate are also confirmed regulators of the activity-dependent development of functional neuronal networks during critical periods of early postnatal life [37,38]. These critical periods represent developmental time windows during which brain circuits that subserve a given function are particularly receptive to acquiring a certain kind of information or even need that instructive signal for their continued development . A delicate balance between the excitatory and inhibitory signals has a key importance in appropriate network development. In addition to drastic pharmacological perturbations of neuronal activity, even small changes in the relative amounts of excitation and inhibition can markedly alter information processing .
Immature neurons and neuronal circuits are particularly sensitive to external stimuli during the synaptogenetic period. While, as discussed above, endogenous GABA and glutamate are clearly key factors guiding CNS morphogenesis, exogenous stimulation or blockade of GABAergic and glutaminergic signalling pathways can also trigger cell death in the developing brain . It was shown over 30 yr ago, that the subcutaneous injection of high doses of glutamate induces acute neuronal necrosis in several brain regions in newborn mice as well as in monkeys [41,42]. In turn, blockade of the NMDA receptor during synaptogenesis triggers widespread apoptosis in the developing brain .
Studies in transgenic mice with deficient GABAergic or glutaminergic signalling pathways further improved our understanding of the role of these molecules during the critical periods of neuronal network development. GABA is synthesized by glutamic acid decarboxylase, produced by two distinct genes Gad65 and Gad67. While deletion of the Gad67gene is lethal and eliminates most cortical GABA content , Gad65-knockout mice are viable but show poor GABA release from synaptic vesicles on stimulation . Importantly, this observed decrease in GABA-mediated synaptic neurotransmission resulted in an impairment of activity dependent refinement of functional connections in the developing visual cortex, underscoring the role of GABAergic neurotransmission in synaptic plasticity .
In transgenic animals with non-functional NMDA receptors due to a lack of the NR1 receptor subunit, there is increased apoptosis in several brain areas during the period of naturally occurring cell death and synaptogenesis  and these animals die soon after birth . The generation of region-specific knockouts where non-functional NMDA receptors are only located in the cerebral cortex allowed us to study the role of this receptor in the formation of cortical circuits during early postnatal life . These experiments revealed the important role of NMDA receptor during the critical period in several cortical systems, such as the formation of the somatosensory barrel cortex  and the shaping of ocular dominance columns in the visual cortex .
The fetal alcohol syndrome is a dramatic clinical demonstration of how exogenous perturbations of GABAergic and glutaminergic signalling can affect brain development. Administration of ethanol, which has both NMDA antagonist and GABA mimetic properties, to pregnant rodents provokes disturbances of cortical lamination, neuronal ectopia and reduced thickness of the cortical mantle in the offspring . In fact, ethanol has been shown to inhibit proliferation of neuronal precursor cells, to impair their migration and to induce neuronal death, which are possibly the neurobiological cause of the reduced brain mass and lifelong neurobehavioural disturbances associated with human fetal alcohol syndrome . In contrast to immature cells, differentiated neurons are less sensitive to the pharmacological modulation of GABAergic and glutaminergic signalling . Altogether, these findings suggest that the increased vulnerability of neurons is mainly confined to the synaptogenetic period, and that once this period is over, transient pharmacological manipulation of neuronal activity will not, or at least less, affect neuronal survival and optimal network function.
Effect of anaesthetics on CNS development
Based on our increasing understanding of the important developmental role of neurotransmitter systems, it is thus not surprising that the past few years have yielded a veritable explosion of publications claiming adverse effects of anaesthetics on the immature brain. In the following section, we will review existing laboratory data, by focusing on drugs that were clearly shown to affect some aspects of CNS development either in vitro or in vivo.
Ketamine is a non-competitive antagonist at the phencyclidine (PCP) site of the NMDA receptor and, by preventing excitotoxic actions of endogenous excitatory amino acids such as glutamate and aspartate, might have a neuroprotective role in ischaemia-induced and seizure-related brain damage [50–52]. The first indications that, in addition to its potential benefits, ketamine can also induce pathological changes in the CNS came from the observations that subcutaneous administration of PCP to adult rats induced cytopathological changes (vacuolization of neuronal cytoplasm) in cerebrocortical neurons . These effects were detectable as early as 2 h after drug exposure and neuronal morphology returned to normal when only one single dose of PCP was applied. In this same study, ketamine mimicked the effect of PCP when administered subcutaneously at 40 mg kg−1, while lower doses did not induce vacuolar neurodegeneration . These observations were further extended to the developing brain in another study, in which a series of seven subcutaneous injections of ketamine (20 mg kg−1 at each dose) spaced evenly over 9 h induced extensive apoptotic neurodegeneration in 7-day-old rat pups .
The relevance of these rodent data has recently been confirmed in primates, where ketamine induced extensive neuronal cell death in the cerebral cortex of rhesus monkey fetuses when the mother was exposed to this anaesthetic for 24 h . The ensemble of these experiments thus strongly supports the contention that long-term exposure to ketamine at anaesthetic concentrations could indeed exert neurotoxic effects on the developing CNS.
Whether short-term ketamine exposure, as frequently used in current children’s anaesthesia practice, can induce cell death in the developing brain is a controversial issue. Several independent observations indicate that, in contrast to repeated injections of ketamine aimed to produce long-term anaesthesia, a single anaesthetic dose of this drug does not induce neuronal apoptosis [53,55,56]. These results were, however, challenged by new experiments demonstrating that even relatively mild exposure to ketamine can trigger apoptosis in the developing mouse brain . In this latter study, using immunocytochemical labelling with a specific antibody against the early apoptotic marker caspase 3, apoptotic cell bodies could be detected in several brain areas as early as 4 h following single subcutaneous injections of ketamine to 7-day-old mice calculated to give an anaesthetic or a subanaesthetic concentration (40 and 10 mg kg−1, respectively). While it is clear that further experiments are needed to elucidate the impact of a single ketamine dose on neuronal responses, these results raise the intriguing possibility that even a brief apoptogenic stimulus might alter neuronal development during the peak synaptogenetic period.
An important point of concern in terms of neurotoxicity is that neuronal apoptosis is not the only parameter to be considered in evaluating potential adverse effects of ketamine or other anaesthetics on neuronal development. It is now well established that interference with the finely tuned molecular mechanisms, guiding the formation of neuronal dendritic arbors in the developing brain, can lead to persistent dysfunction of the CNS . Thus, understanding whether anaesthetics modify dendritic arbor expansion during CNS development is of utmost interest. To investigate this issue, we have recently developed an in vitro model system, in which immature neuroblasts are isolated from the subventricular zone of newborn rats . This purified cell population develops into GABAergic interneurons in low-density cultures with a nice dendritic arbor. Based on previous experimental and clinical studies measuring ketamine plasma concentrations following a single shot or repeated administration in rodents and human beings [56,60,61], we have recently explored the dose- and exposure time-dependent effects of ketamine on the differentiation and survival of GABAergic neurons in these cultures . We found that ketamine, but not the non-competitive NMDA receptor antagonist MK 801, rapidly induced apoptosis of developing neurons when administered at concentrations previously reported to induce cell death in vivo (≥10 μg mL−1) . Neither survival nor long-term dendritic development was altered when differentiating neurons were exposed to lower, subanaesthetic concentrations (≤2 μg mL−1) of this anaesthetic for up to 8 h. In contrast, long-term exposure (>24 h) of neurons to ketamine at concentrations as low as 0.01 μg mL−1 severely impaired dendritic arbor development. These new data suggest that long-term use of even low concentrations of ketamine, such as an adjuvant to postoperative sedation and pain control, could potentially interfere with the dendritic development of immature neurons.
While ketamine is primarily considered to exert its effects by blocking NMDA receptors, we found substantial differences between this anaesthetic and another non-competitive NMDA receptor antagonist, MK 801, on neuronal differentiation and survival . The fact that a 1-h treatment with ketamine but not with MK 801 was sufficient to trigger an important apoptosis of GABAergic neurons raises the possibility that, at high dose regimens, ketamine-induced neurotoxicity is, at least partially, independent of NMDA receptor blockade. These data were further confirmed by experiments showing that exposure of cultures to MK 801 for up to 4 h affect neither the survival nor differentiation of developing neurons. One plausible explanation of these observations would be that, in addition to NMDA receptor blockade, ketamine also interacts with a multitude of signalling pathways mediating neurotransmission in the CNS . Indeed, ketamine induces release of dopamine, serotonine and norepinephrine in the brain [64,65] and recent experimental evidence indicates that this anaesthetic also interferes with the reuptake of these amines from the extracellular space by inhibiting monoamine transporters [65,66].
It is of interest that accumulation of monoamines has been reported to trigger extensive neurodegeneration in rodents  and blockade of the serotonin transporter has been shown to reduce the complexity of dendritic arbor architecture of hippocampal pyramidal neurons . Ketamine also induces the release of adenosine from nerve terminals  and there is now evidence that adenosine A2A receptors play a permissive role in the metabotropic glutamate receptor-mediated potentiation of NMDA signalling . It is thus possible that, in the presence of higher concentrations of ketamine, additive or synergistic effects between these molecular mechanisms and signalling pathways could rapidly initiate dendritic remodeling and/or apoptosis. Alternatively, large doses of ketamine could exert a non-specific neurotoxic effect.
Little is known about the potentially adverse effects of ketamine on neuronal progenitor proliferation and cell migration. Recent observations indicate a possible role for ketamine in progenitor proliferation in postnatal neurogenic zones. In adult rats given subanaesthetic concentrations of ketamine for 5 consecutive days, the proliferation marker bromodeoxyuridine revealed enhanced neurogenesis in the hippocampal subgranular zone . To our knowledge, no other studies exist on the effect ketamine exposure has on the immature brain before the synaptogenetic period. Given the important role of these earlier developmental stages in the proper formation of the CNS, further research is needed to clarify these issues.
Propofol (2,6-diisopropyl phenol) is an alkyl phenol derivative dissolved in a lipid emulsion. This drug potentiates the effect of GABA in the CNS by inducing tyrosine kinase-mediated phosphorylation of the β subunits of the GABAA receptor complex . Although there is controversy regarding the use of propofol in children [73,74], this agent is commonly administered to young children, including neonates [75,76]. Selective toxicity of propofol for GABAergic neurons but not for astroglial cells has been shown in aggregated cell cultures of the fetal rat telencephalon . Accordingly, in two recent studies, we provided morphological evidence on the potentially deleterious effects of propofol on neuronal differentiation and survival. Using primary cultures from the newborn rat cerebral cortex, we showed that propofol induced a dose-dependent loss of developing GABAergic neurons while the survival of other cell types, such as oligodendrocytes and astrocytes, was not affected .
To further elucidate the effects of low, but clinically relevant concentrations of propofol on dendritic arbor development, we took advantage of our previously described in vitro model of subventricular zone-derived GABAergic interneurons . Using this model, we showed that propofol induced death of GABAergic neurons at concentrations of 50 μg mL−1 or greater. While propofol did not trigger cell death at lower concentrations, as little as 1 μg mL−1 of this drug significantly altered several aspects of dendritic development, and exposure as short as 4 h resulted in a persistent decrease in dendritic growth . In contrast to differentiating neurons, we found no evidence for the neurotoxic effects of propofol in 3-week-old organotypic slice cultures of the hippocampus . In line with these data, in a series of preliminary experiments we observed no adverse effects of even high concentrations (>50 μg mL−1) of propofol when this drug was applied to 3-week-old primary dissociated cortical cultures containing neurons with a highly differentiated dendritic arbor. It is thus possible that potential neurotoxic effects of propofol depend on the developmental stage at which neurons are exposed to it. These results bring further arguments in favour of the assumption that immature neurons are particularly sensitive to anaesthetics .
Similar to propofol, midazolam also acts at the GABAA receptor complex and is widely used in children, for both anaesthetic and sedative purposes . In our recent in vitro study, we found that the effects of midazolam on differentiating GABAergic neurons are rather different from those of propofol . While, as described above, low concentrations of propofol already had a negative impact on dendritic development, even high concentrations of midazolam (>25 μg mL−1) did not have any effect on neuronal differentiation and survival in this model. The reason for this difference is unclear but might be explained by the different sites of action of these molecules on the GABAA receptor complex. It is now well established that the profile of actions of drugs that modulate GABAergic activity are largely dependent upon the receptor subunit by which they interact , and whereas propofol induces tyrosine kinase-mediated phosphorylation on the β subunits of the GABAA receptor , benzodiazepines attach selectively to the α subunits .
Another plausible explication for these differential effects between the two drugs comes from the increasing number of observations indicating that propofol can also act via GABAA receptor-independent mechanisms. Indeed, this drug has been shown to inhibit NMDA receptors in hippocampal neurons  as well as interfere with signalling through nicotinic acetylcholine receptors  and T-type Ca2+ channels . Finally, it has recently been demonstrated that propofol but not midazolam induces the phosphorylation of actin [86,87]. Thus, phosphorylation-induced reorganization of the actin cytoskeleton in developing GABAergic neurons seen with propofol but not with midazolam could also provide an alternative explanation for this differential effect.
Recent observations suggest that midazolam can also induce apoptosis in the immature CNS . In these experiments, 7-day-old mice were given a single, low, subcutaneous dose of midazolam (9 mg kg−1). The authors demonstrated that this treatment, although insufficient to produce anaesthesia, induced a significant neuroapoptotic response in the cerebral cortex as well as in the basal ganglia. While interspecies differences (mice vs. rats) cannot be fully excluded to account for these apparent discrepancies between these and our own midazolam results, we believe that differences in neuronal sub-types detected and evaluated in these two studies would be a more plausible explanation. In fact, in our model, we examined the effect of midazolam in a rather homogenous population of GABAergic neurons isolated from the subventricular zone. In contrast, Young and colleagues observed midazolam-induced apoptosis in several brain areas, but the sub-type of dying neurons has not been identified . As the differential sensitivity of neuronal sub-classes to apoptotic stimuli has been well established, it is thus possible that midazolam might induce cell death in subpopulations of neurons different from those studied in our model.
While the exact mechanisms by which volatile agents induce anaesthesia remain to be determined, all have GABA mimetic and/or NMDA antagonist properties. Halothane was amongst the first anaesthetics reported to alter brain development more than 20 yr ago. When rats were chronically exposed to halothane anaesthesia during the entire gestation period, dendritic length and branching as well as cerebral synaptic density were severely impaired [88,89]. The effect of halothane on dendritic growth appeared to be enduring, and the delay in initial dendritic growth caused by halothane was not compensated for by an increased rate of dendritic growth once the drug was withdrawn. Importantly, the effect of halothane treatment on dendritic growth was associated with learning impairment, decreased exploratory behaviour and decreased nociceptive reactivity .
Isoflurane 1.5% has recently been reported to induce neuronal degeneration in organotypic hippocampal slice cultures . In these experiments, isoflurane-induced cell death occurred in cultures obtained from postnatal 7-day-old rats (PND7) but not in those from PND4 or PND14 rat pups. Moreover, this effect was only observed with an isoflurane exposure of at least 5 h. This in vitro study further supports the notion of both age- and duration-dependent relationship between anaesthesia administration and perinatal neuronal death. Interesting new data show that even shorter exposure time (4 h) to isoflurane in neonatal rat pups, although not affecting survival, can induce a significant decrease in hippocampal stem cell proliferation, leading to a long-term impairment of neurocognitive function revealed by the fear conditioning test .
Nitrous oxide is a potent antagonist of the NMDA type of glutamate receptor. When timed pregnant rats were exposed to 75% N2O/25% O2 mixtures on gestational days 14 and 15 for 8 h day−1, permanently altered behavioural deficits could be detected in offsprings without any accompanying physical abnormalities . In line with the results observed following ketamine administration, there is now evidence that a 3-h-long exposure to N2O can also trigger apoptosis in the developing brain .
Combined use of anaesthetics
During the perioperative period, patients are usually exposed to a combination of different anaesthetics either simultaneously or successively. As the majority of these drugs have GABA mimetic and/or NMDA antagonist properties, the question whether the combined use of anaesthetics could act additively or synergistically in terms of neurotoxicity is of utmost importance. While one major argument to provide multimodal anaesthesia is to avoid the potential side-effects associated with higher concentrations of the individual drugs, recent laboratory results would suggest the need for a risk/benefit re-evaluation of this practice. Indeed, a large body of experimental evidence suggests that concurrent use of several anaesthetics can potentiate cerebrocortical damage. Co-administration of even low concentrations of ketamine and nitrous oxide enhances the neurotoxic reaction to a much greater degree that can be explained by simple additivity between these agents . Recently, co-administration of sedative concentrations of midazolam and ketamine to the infant mouse brain has been shown to be more effective in inducing apoptosis than either of these drugs alone . Importantly, exposure of 7-day-old rats to a combined midazolam–nitrous oxide–isoflurane anaesthesia for 6 h led to widespread neurodegeneration in the developing brain, and this was accompanied by persistent learning deficits .
Interesting new data from this same research group provide some insights into the molecular mechanisms of anaesthesia-induced activation of apoptotic pathways in immature neurons . The abovementioned anaesthesia protocol, using midazolam–nitrous oxide–isoflurane, rapidly induced significant changes in the expression pattern of brain-derived neurotrophic factor (BDNF) in the brain of rat pups. BDNF is a member of the neurotrophin family of growth factors and is implicated in neuronal survival, differentiation and synaptic plasticity, via activation of its high-affinity receptor TrkB . This neurotrophin also binds to the so-called low-affinity neurotrophin receptor p75NTR, leading to the activation of apoptotic cascades . In the cerebral cortex of 7-day-old rats, as little as a 2-h exposure to anaesthesia significantly increased the amounts of BDNF. This was accompanied by an increase in the expression of the p75NTR receptor. In contrast, anaesthesia induced an important decrease in BDNF levels without affecting p75NTR receptor expression in the thalamus, leading to a reduced BDNF-dependent activation of the TrkB receptor. Altogether, these results suggest that anaesthesia can disrupt CNS development by interacting with the multifaceted molecular pathways of neurotrophin signalling.
Deciphering the molecular pathways involved in anaesthesia-induced adverse effects on the developing CNS would allow us to develop therapeutic strategies to prevent these unwanted complications. In this context, administration of β-estradiol has been shown to reduce anaesthesia-induced apoptosis in the developing brain . This sexual hormone is known to activate the Akt serine/threonine kinase, which, in turn, is an important factor for neuronal survival pathways. The apoptotic neurodegeneration induced by the midazolam–nitrous oxide–isoflurane ‘cocktail’ has been recently reported to be dose-dependently reduced by the co-administration of melatonin that, in addition to its sleep-promoting activity, also exerts antioxidant effects by improving mitochondrial homeostasis and stabilizing the inner mitochondrial membrane . These encouraging observations open a whole new line of research aimed at counteracting anaesthesia-induced neurotoxicity and further studies should be conducted to address this important issue.
Extrapolation of laboratory results to clinical practice
It is of course extremely difficult to evaluate the clinical relevance of experimental observations that suggest anaesthesia-related neurotoxicity in the developing brain. A first important question concerns the possibility of interspecies differences in terms of drug effects . In this context, it is, however, important to note that, in addition to rodents, anaesthetic and subanaesthetic doses of currently used anaesthetics have now been shown to induce apoptosis in other species such as pigs and monkeys [54,100]. Another essential criticism concerning the significance of animal experiments for human anaesthetic practice is the relatively long exposure time needed to produce detectable neurotoxic effects in the majority of laboratory studies . In fact, from a developmental perspective, several hour long exposures to anaesthetics in rodents would be equivalent to producing general anaesthesia for days or even weeks in the human neonate .
However, recent results contradict these arguments to some extent, since they show that even a single exposure to subanaesthetic doses of anaesthetics could trigger two- to fourfold increases in neuronal apoptosis in the mouse brain during the synaptogenetic period . Also, as discussed previously, in vitro data indicate that short-term exposure to anaesthetics can also impair neuronal development by interfering with dendritic growth and branching without inducing cell death [62,79]. Given the utmost importance of neuronal dendritic architecture in appropriate information processing, one essential next step will be to determine how neuronal dendritic arbor development is influenced by anaesthetics in a more complex and physiological environment, using organotypic slice cultures and in vivo animal experiments. These experiments, combined with long-term assessment of behavioural outcome following short-term anaesthesia, would probably help us to better understand the impact of anaesthetics on CNS development.
Differences in the concentrations of drugs required to produce anaesthesia in different species further complicate the issue of anaesthesia-induced developmental neurotoxicity. For example, subanaesthetic plasma concentrations of ketamine in human beings are around 0.1–0.5 μg mL−1 [103,104], while i.v. administration to children at doses of 3 mg kg−1 to induce anaesthesia was associated with blood levels of 1–2 μg mL−1 [60,61]. While no direct comparison in terms of plasma concentrations exists with rodents, as much as 40 mg kg−1 of subcutaneously injected ketamine is insufficient to produce anaesthesia in young mice . Furthermore, recent observations indicate that plasma levels of ketamine are around 6 μg mL−1 following a single subcutaneous injection of 20 mg kg−1 . Altogether, these studies strongly suggest that effective plasma concentrations, and probably ‘on-site’ brain concentrations as well, of ketamine needed to produce anaesthesia is significantly higher in rodents than in human beings, raising further difficulties in the extrapolation of these experiments to human infants.
Finally, one can argue that experimental conditions in these animal experiments are very much different from those associated with surgical anaesthesia and complex perioperative management . First, based on the neuronal stimulation hypothesis , preoperative stress and painful stimuli during surgery can activate NMDA and other excitatory receptors in the immature brain, and anaesthetic drugs could thus reduce extreme degrees of neuronal excitation . This clinical situation is in contrast to experimental settings where anaesthesia was administered without painful stimuli and, consequently, the effect of anaesthetics on the suppression of basal neural activity is evaluated. Second, human neonates and children routinely receive nutritional support and metabolic monitoring in the perioperative period, thus minimizing the risk of hypoglycaemia and impaired nutrition. In contrast, although this issue is controversial , rodent pups do not suckle well after general anaesthesia, resulting in a prolonged decrease in weight gain compared to non-anaesthetized litter-mates . Given that the role of malnutrition in decreased brain growth and learning disabilities is well established [108,109], one cannot fully exclude the possibility that neurotoxic effects of anaesthetics are, at least partially, related to impaired nutrition in the perioperative period in animal studies.
Anaesthesia-induced neurotoxicity is a highly debated and controversial issue [101,107,110]. Based on the developmental role of GABAergic and glutaminergic neurotransmission in shaping CNS development, there is no doubt that experimental as well as clinical research in this domain is of primary importance. Indeed, every year, a large number of human fetuses and neonates receive anaesthesia worldwide  and perioperative morbidity and mortality is approximately tenfold higher in neonates than children in older age groups . Despite the fact that animal studies provide, at best, an imprecise basis for evaluating human risk, such experiments still remain very useful. In fact, a clearer understanding of the highly complex molecular and cellular mechanisms by which anaesthesia can alter survival or development of immature neurons and glia cells during brain development, particularly in the synaptogenetic period, could also contribute to the development of therapeutic strategies for preventing long-term neurobehavioral abnormalities in human beings. Animal experiments are of utmost importance in elucidating fundamental and, as yet, unanswered issues regarding the effects of anaesthetics on neuronal networks before, during and after the critical developmental period of a given brain region. Indeed, the question of whether a transient short-term exposure to anaesthetics at defined stages of CNS development can initiate permanent morpho-functional changes in the brain is still open. Finally, as it would be practically and ethically impossible to establish a dose- and exposure-time response curve of anaesthesia-induced neurotoxicity in human infants, such experiments should be performed in non-human primates to further elucidate this question.
Multiple lines of evidence support the necessity of adequate anaesthesia in neonates and young children undergoing surgery [2–7]. It is thus definitely clear that, in spite of the wealth of observations presented in this review and related to anaesthesia-induced toxicity, clinicians should not withhold anaesthesia from newborns. On the other hand, the results of these laboratory experiments should provide further arguments to promote clinical research on this topic. Despite the difficulty in conducting clinical trials in newborn and children’s populations to assess the effect of anaesthetics on CNS development, these future studies will be necessary to bridge the gap between laboratory neuroscience and clinical medicine [110,113].
The authors thank Professors E. Tassonyi, A. Kato and M. Tramer for their helpful comments during the preparation of this manuscript.
1. Hobbs AJ, Bush GH, Downham DY. Peri-operative dreaming and awareness in children. Anaesthesia
1988; 43: 560–562.
2. Anand KS. Relationships between stress responses and clinical outcome in newborns, infants, and children. Crit Care Med
1993; 21: S358–S359.
3. Bouwmeester NJ, Anand KJ, van Dijk M et al.
Hormonal and metabolic stress responses after major surgery in children aged 0–3 years: a double-blind, randomized trial comparing the effects of continuous versus intermittent morphine. Br J Anaesth
2001; 87: 390–399.
4. van Lingen RA, Simons SH, Anderson BJ, Tibboel D. The effects of analgesia in the vulnerable infant during the perinatal period. Clin Perinatol
2002; 29: 511–534.
5. Taddio A, Shah V, Gilbert-MacLeod C, Katz J. Conditioning and hyperalgesia in newborns exposed to repeated heel lances. JAMA
2002; 288: 857–861.
6. Tobiansky R, Lui K, Roberts S, Veddovi M. Neurodevelopmental outcome in very low birthweight infants with necrotizing enterocolitis requiring surgery. J Paediatr Child Health
1995; 31: 233–236.
7. Chacko J, Ford WD, Haslam R. Growth and neuro-developmental outcome in extremely-low-birth-weight infants after laparotomy. Pediatr Surg Int
1999; 15: 496–499.
8. Represa A, Ben-Ari Y. Trophic actions of GABA on neuronal development. Trends Neurosci
2005; 28: 278–283.
9. Waters KA, Machaalani R. NMDA receptors in the developing brain and effects of noxious insults. Neurosignals
2004; 13: 162–174.
10. Nguyen L, Rigo JM, Rocher V et al.
Neurotransmitters as early signals for central nervous system development. Cell Tissue Res
2001; 305: 187–202.
11. Herlenius E, Lagercrantz H. Development of neurotransmitter systems during critical periods. Exp Neurol
2004; 190(Suppl 1): S8–S21.
12. Benitez-Diaz P, Miranda-Contreras L, Mendoza-Briceno RV et al.
Prenatal and postnatal contents of amino acid neurotransmitters in mouse parietal cortex. Dev Neurosci
2003; 25: 366–374.
13. Lujan R, Shigemoto R, Lopez-Bendito G. Glutamate
and GABA receptor signalling in the developing brain. Neuroscience
2005; 130: 567–580.
14. McMahon D. Chemical messengers in development: a hypothesis. Science
1974; 185: 1012–1021.
15. Owens DF, Kriegstein AR. Developmental neurotransmitters? Neuron
2002; 36: 989–991.
16. The Boulder Committee. Embryonic vertebrate central nervous system: revised terminology. Anat Rec
1970; 166: 257–261.
17. Sidman RL, Miale IL, Feder N. Cell proliferation and migration in the primitive ependymal zone: an autoradiographic study of histogenesis in the nervous system. Exp Neurol
1959; 1: 322–333.
18. Altman J. Autoradiographic and histological studies of postnatal neurogenesis. IV. Cell proliferation and migration in the anterior forebrain, with special reference to persisting neurogenesis in the olfactory bulb. J Comp Neurol
1969; 137: 433–457.
19. Doetsch F, Caille I, Lim DA et al.
Subventricular zone astrocytes are neural stem cells in the adult mammalian brain. Cell
1999; 97: 703–716.
20. LoTurco JJ, Owens DF, Heath MJ et al.
GABA and glutamate
depolarize cortical progenitor cells and inhibit DNA synthesis. Neuron
1995; 15: 1287–1298.
21. Antonopoulos J, Pappas IS, Parnavelas JG. Activation of the GABAA
receptor inhibits the proliferative effects of bFGF in cortical progenitor cells. Eur J Neurosci
1997; 9: 291–298.
22. Behar TN, Li YX, Tran HT et al.
GABA stimulates chemotaxis and chemokinesis of embryonic cortical neurons via calcium-dependent mechanisms. J Neurosci
1996; 16: 1808–1818.
23. Behar TN, Schaffner AE, Scott CA et al.
GABA receptor antagonists modulate postmitotic cell migration in slice cultures of embryonic rat cortex. Cereb Cortex
2000; 10: 899–909.
24. Haydar TF, Wang F, Schwartz ML, Rakic P. Differential modulation of proliferation in the neocortical ventricular and subventricular zones. J Neurosci
2000; 20: 5764–5774.
25. Kornack DR, Rakic P. Changes in cell-cycle kinetics during the development and evolution of primate neocortex. Proc Natl Acad Sci USA
1998; 95: 1242–1246.
26. Liu X, Wang Q, Haydar TF, Bordey A. Nonsynaptic GABA signalling in postnatal subventricular zone controls proliferation of GFAP-expressing progenitors. Nat Neurosci
2005; 8: 1179–1187.
27. Behar TN, Scott CA, Greene CL et al. Glutamate
acting at NMDA receptors stimulates embryonic cortical neuronal migration. J Neurosci
1999; 19: 4449–4461.
28. Rivera C, Voipio J, Kaila K. Two developmental switches in GABAergic signalling: the K+
cotransporter KCC2 and carbonic anhydrase CAVII. J Physiol
2005; 562: 27–36.
29. Miller M. Maturation of rat visual cortex. I. A quantitative study of Golgi-impregnated pyramidal neurons. J Neurocytol
1981; 10: 859–878.
30. Miller M, Peters A. Maturation of rat visual cortex. II. A combined Golgi-electron microscope study of pyramidal neurons. J Comp Neurol
1981; 203: 555–573.
31. McAllister AK. Cellular and molecular mechanisms of dendrite growth. Cereb Cortex
2000; 10: 963–973.
32. Cline HT. Dendritic arbor development and synaptogenesis. Curr Opin Neurobiol
2001; 11: 118–126.
33. Wong RO, Ghosh A. Activity-dependent regulation of dendritic growth and patterning. Nat Rev Neurosci
2002; 3: 803–812.
34. Chen Y, Ghosh A. Regulation of dendritic development by neuronal activity. J Neurobiol
2005; 64: 4–10.
35. Jan YN, Jan LY. The control of dendrite development. Neuron
2003; 40: 229–242.
36. Rajan I, Cline HT. Glutamate
receptor activity is required for normal development of tectal cell dendrites in vivo
. J Neurosci
1998; 18: 7836–7846.
37. Hensch TK. Critical period regulation. Annu Rev Neurosci
2004; 27: 549–579.
38. Hensch TK. Critical period plasticity in local cortical circuits. Nat Rev Neurosci
2005; 6: 877–888.
39. Liu G. Local structural balance and functional interaction of excitatory and inhibitory synapses in hippocampal dendrites
. Nat Neurosci
2004; 7: 373–379.
40. Olney JW. New insights and new issues in developmental neurotoxicology. Neurotoxicology
2002; 23: 659–668.
41. Olney JW. Brain lesions, obesity, and other disturbances in mice treated with monosodium glutamate
1969; 164: 719–721.
42. Olney JW, Sharpe LG. Brain lesions in an infant rhesus monkey treated with monsodium glutamate
1969; 166: 386–388.
43. Ikonomidou C, Bosch F, Miksa M et al.
Blockade of NMDA receptors and apoptotic neurodegeneration in the developing brain. Science
1999; 283: 70–74.
44. Asada H, Kawamura Y, Maruyama K et al.
Cleft palate and decreased brain gamma-aminobutyric acid
in mice lacking the 67-kDa isoform of glutamic acid decarboxylase. Proc Natl Acad Sci USA
1997; 94: 6496–6499.
45. Hensch TK, Fagiolini M, Mataga N et al.
Local GABA circuit control of experience-dependent plasticity in developing visual cortex. Science
1998; 282: 1504–1508.
46. Adams SM, de Rivero Vaccari JC, Corriveau RA. Pronounced cell death in the absence of NMDA receptors in the developing somatosensory thalamus. J Neurosci
2004; 24: 9441–9450.
47. Li Y, Erzurumlu RS, Chen C et al.
Whisker-related neuronal patterns fail to develop in the trigeminal brainstem nuclei of NMDAR1 knockout mice. Cell
1994; 76: 427–437.
48. Iwasato T, Datwani A, Wolf AM et al.
Cortex-restricted disruption of NMDAR1 impairs neuronal patterns in the barrel cortex. Nature
2000; 406: 726–731.
49. Ikonomidou C, Bittigau P, Ishimaru MJ et al.
Ethanol-induced apoptotic neurodegeneration and fetal alcohol syndrome. Science
2000; 287: 1056–1060.
50. Albers GW, Goldberg MP, Choi DW. N
-methyl-d-aspartate antagonists: ready for clinical trial in brain ischemia? Ann Neurol
1989; 25: 398–403.
51. Bullock R. Strategies for neuroprotection with glutamate
antagonists. Extrapolating from evidence taken from the first stroke and head injury studies. Ann NY Acad Sci
1995; 765: 272–278discussion 298.
52. Himmelseher S, Durieux ME. Revising a dogma: ketamine for patients with neurological injury? Anesth Analg
2005; 101: 524–534.
53. Olney JW, Labruyere J, Price MT. Pathological changes induced in cerebrocortical neurons by phencyclidine and related drugs. Science
1989; 244: 1360–1362.
54. Scallet AC, Divine R, Wang C et al.
in prenatal rhesus monkeys: distribution of neuronal damage. Soc Neurosci Abst
55. Hayashi H, Dikkes P, Soriano SG. Repeated administration of ketamine may lead to neuronal degeneration in the developing rat brain. Paediatr Anaesth
2002; 12: 770–774.
56. Scallet AC, Schmued LC, Slikker W et al.
of ketamine: morphometric confirmation, exposure parameters, and multiple fluorescent labeling of apoptotic neurons. Toxicol Sci
2004; 81: 364–370.
57. Young C, Jevtovic-Todorovic V, Qin YQ et al.
Potential of ketamine and midazolam, individually or in combination, to induce apoptotic neurodegeneration in the infant mouse brain. Br J Pharmacol
2005; 146: 189–197.
58. Webb SJ, Monk CS, Nelson CA. Mechanisms of postnatal neurobiological development: implications for human development. Dev Neuropsychol
2001; 19: 147–171.
59. Gascon E, Vutskits L, Zhang H et al.
Sequential activation of p75 and TrkB is involved in dendritic development of subventricular zone-derived neuronal progenitors in vitro
. Eur J Neurosci
2005; 21: 69–80.
60. Malinovsky JM, Servin F, Cozian A et al.
Ketamine and norketamine plasma concentrations after i.v., nasal and rectal administration in children. Br J Anaesth
1996; 77: 203–207.
61. Weber F, Wulf H, Gruber M, Biallas R. S
-ketamine and S
-norketamine plasma concentrations after nasal and i.v. administration in anesthetized children. Paediatr Anaesth
2004; 14: 983–988.
62. Vutskits L, Gascon E, Tassonyi E, Kiss JZ. Effect of ketamine on dendritic arbor development and survival of immature GABAergic neurons in vitro
. Toxicol Sci
2006; e-pub ahead.
63. Adams HA. Mechanisms of action of ketamine. Anaesthesiol Reanim
1998; 23: 60–63.
64. Kari HP, Davidson PP, Kohl HH, Kochhar MM. Effects of ketamine on brain monoamine levels in rats. Res Commun Chem Pathol Pharmacol
1978; 20: 475–488.
65. Tso MM, Blatchford KL, Callado LF et al.
Stereoselective effects of ketamine on dopamine, serotonin and noradrenaline release and uptake in rat brain slices. Neurochem Int
2004; 44: 1–7.
66. Nishimura M, Sato K, Okada T et al.
Ketamine inhibits monoamine transporters expressed in human embryonic kidney 293 cells. Anesthesiology
1998; 88: 768–774.
67. Bozzi Y, Borrelli E. Dopamine in neurotoxicity
and neuroprotection: what do D(2) receptors have to do with it? Trends Neurosci
2006; 29: 167–174.
68. McKittrick CR, Magarinos AM, Blanchard DC et al.
Chronic social stress reduces dendritic arbors in CA3 of hippocampus and decreases binding to serotonin transporter sites. Synapse
2000; 36: 85–94.
69. Mazar J, Rogachev B, Shaked G et al.
Involvement of adenosine in the antiinflammatory action of ketamine. Anesthesiology
2005; 102: 1174–1181.
70. Tebano MT, Martire A, Rebola N et al.
Adenosine A2A receptors and metabotropic glutamate
5 receptors are co-localized and functionally interact in the hippocampus: a possible key mechanism in the modulation of N
-methyl-d-aspartate effects. J Neurochem
2005; 95: 1188–1200.
71. Keilhoff G, Bernstein HG, Becker A et al.
Increased neurogenesis in a rat ketamine model of schizophrenia. Biol Psychiatry
2004; 56: 317–322.
72. Bjönström K, Sjölander A, Schippert A, Eintrei C. A tyrosine kinase regulates propofol-induced modulation of the β-subunit of the GABAA
receptor and release of intracellular calcium in cortical rat neurons. Acta Physiol Scand
2002; 175: 227–235.
73. Wooltorton E. Propofol contraindicated for sedation of pediatric intensive care patients. CMAJ
2002; 167: 507.
74. Crawford MW, Dodgson BG, Holtby HH, Roy WL. Propofol syndrome in children. CMAJ
2003; 168: 669author reply 669–70.
75. Westrin P. The induction dose of propofol in infants 1–6 months of age and in children 10–16 years of age. Anesthesiology
1991; 74: 455–458.
76. Murat I, Billard V, Vernois J et al.
Pharmacokinetics of propofol after a single dose in children aged 1–3 years with minor burns. Comparison of three data analysis approaches. Anesthesiology
1996; 84: 526–532.
77. Honegger P, Matthieu JM. Selective toxicity of the general anesthetic propofol for GABAergic neurons in rat brain cell cultures. J Neurosci Res
1996; 45: 631–636.
78. Spahr-Schopfer I, Vutskits L, Toni N et al.
Differential neurotoxic effects of propofol on dissociated cortical cells and organotypic hippocampal cultures. Anesthesiology
2000; 92: 1408–1417.
79. Vutskits L, Gascon E, Tassonyi E, Kiss JZ. Clinically relevant concentrations of propofol but not midazolam alter in vitro
dendritic development of isolated gamma-aminobutyric acid
-positive interneurons. Anesthesiology
2005; 102: 970–976.
80. Arcangeli A, Antonelli M, Mignani V, Sandroni C. Sedation in PACU: the role of benzodiazepines. Curr Drug Targets
2005; 6: 745–748.
81. Mohler H, Fritschy JM, Vogt K et al.
Pathophysiology and pharmacology of GABAA
receptors. Handb Exp Pharmacol
2005; 169: 225–247.
82. Stephenson FA, Duggan MJ, Pollard S. The gamma 2 subunit is an integral component of the gamma-aminobutyric acidA
receptor but the alpha 1 polypeptide is the principal site of the agonist benzodiazepine photoaffinity labeling reaction. J Biol Chem
1990; 265: 21160–21165.
83. Orser BA, Bertlik M, Wang LY, MacDonald JF. Inhibition by propofol (2,6 diisopropylphenol) of the N
-methyl-d-aspartate subtype of glutamate
receptor in cultured hippocampal neurones. Br J Pharmacol
1995; 116: 1761–1768.
84. Flood P, Ramirez-Latorre J, Role L. Alpha 4 beta 2 neuronal nicotinic acetylcholine receptors in the central nervous system are inhibited by isoflurane and propofol, but alpha 7-type nicotinic acetylcholine receptors are unaffected. Anesthesiology
1997; 86: 859–865.
85. Todorovic SM, Lingle CJ. Pharmacological properties of T-type Ca2+
current in adult rat sensory neurons: effects of anticonvulsant and anesthetic agents. J Neurophysiol
1998; 79: 240–252.
86. Oscarsson A, Massoumi R, Sjolander A, Eintrei C. Reorganization of actin in neurons after propofol exposure. Acta Anaesthesiol Scand
2001; 45: 1215–1220.
87. Bjornstrom K, Eintrei C. The difference between sleep and anaesthesia
is in the intracellular signal: propofol and GABA use different subtypes of the GABAA
receptor beta subunit and vary in their interaction with actin. Acta Anaesthesiol Scand
2003; 47: 157–164.
88. Uemura E, Bowman RE. Effects of halothane on cerebral synaptic density. Exp Neurol
1980; 69: 135–142.
89. Uemura E, Levin ED, Bowman RE. Effects of halothane on synaptogenesis and learning behavior in rats. Exp Neurol
1985; 89: 520–529.
90. Levin ED, Uemura E, Bowman RE. Neurobehavioral toxicology of halothane in rats. Neurotoxicol Teratol
1991; 13: 461–470.
91. Wise-Faberowski L, Zhang H, Ing R et al.
Isoflurane-induced neuronal degeneration: an evaluation in organotypic hippocampal slice cultures. Anesth Analg
2005; 101: 651–657.
92. Stratmann G, Bickler P, Ku B et al.
Isoflurane increases stem cell proliferation in adult but not in rat neonatal hippocampi. Soc Neurosci Abst
93. Mullenix PJ, Moore PA, Tassinari MS. Behavioral toxicity of nitrous oxide in rats following prenatal exposure. Toxicol Ind Health
1986; 2: 273–287.
94. Jevtovic-Todorovic V, Benshoff N, Olney JW. Ketamine potentiates cerebrocortical damage induced by the common anaesthetic agent nitrous oxide in adult rats. Br J Pharmacol
2000; 130: 1692–1698.
95. Jevtovic-Todorovic V, Hartman RE, Izumi Y et al.
Early exposure to common anesthetic agents causes widespread neurodegeneration in the developing rat brain and persistent learning deficits. J Neurosci
2003; 23: 876–882.
96. Lu LX, Yon JH, Carter LB, Jevtovic-Todorovic V. General anesthesia activates BDNF-dependent neuroapotosis in the developing brain. Apoptosis
2006; May30 (Epub ahead of print).
97. Tenq KK, Hempstead BL. Neurotrophins and their receptors: signaling trios in complex biological systems. Cell Mol Life Sci
2004; 61: 35–48.
98. Yon JH, Carter LB, Reiter RJ, Jevtovic-Todorovic V. Melatonin reduces the severity of anesthesia-induced apoptotic neurodegeneration in the developing rat brain. Neurobiol Dis
2006; 21: 522–530.
99. Berde C, Cairns B. Developmental pharmacology across species: promise and problems. Anesth Analg
2000; 91: 1–5.
100. Rizzi S, Carter LB, Jevtovic-Todorovic V. Clinically used general anesthetics induce neuroapoptosis in the developing piglet brain. Soc Neurosci Abst
101. Anand KJ, Soriano SG. Anesthetic agents and the immature brain: are these toxic or therapeutic? Anesthesiology
2004; 101: 527–530.
102. Clancy B, Darlington RB, Finlay BL. Translating developmental time across mammalian species. Neuroscience
2001; 105: 7–17.
103. Roytblat L, Talmor D, Rachinsky M et al.
Ketamine attenuates the interleukin-6 response after cardiopulmonary bypass. Anesth Analg
1998; 87: 266–271.
104. Zilberstein G, Levy R, Rachinsky M et al.
Ketamine attenuates neutrophil activation after cardiopulmonary bypass. Anesth Analg
2002; 95: 531–536.
105. Lipton SA, Nakanishi N. Shakespeare in love – with NMDA receptors? Nat Med
1999; 5: 270–271.
106. Bhutta AT, Anand KJ. Vulnerability of the developing brain. Neuronal mechanisms. Clin Perinatol
2002; 29: 357–372.
107. Olney JW, Young C, Wozniak DF et al.
Anesthesia-induced developmental neuroapoptosis. Does it happen in humans? Anesthesiology
2004; 101: 273–275.
108. Dobbing J. Undernutrition and the developing brain. The relevance of animal models to the human problem. Am J Dis Child
1970; 120: 411–415.
109. Lucas A, Morley R, Cole TJ. Randomised trial of early diet in preterm babies and later intelligence quotient. BMJ
1998; 317: 1481–1487.
110. Todd MM. Anesthetic neurotoxicity
: the collision between laboratory neuroscience and clinical medicine. Anesthesiology
2004; 101: 272–273.
111. AHRQ:HCUPnet: Healthcare Cost and Utilization Project, Rockville, MD, Agency for Healthcare Research and Quality, 2001 (www.ahrq.gov/hcupnet
112. Cohen MM, Cameron CB, Duncan PG. Pediatric anesthesia morbidity and mortality in the perioperative period. Anesth Analg
1990; 70: 160–167.
113. Anand KJ, Aranda JV, Berde CB et al.
Analgesia and anesthesia for neonates: study design and ethical issues. Clin Ther
2005; 27: 814–843.