How does the brain restore consciousness and cognition after general anesthesia?

Millions of surgical procedures performed each year would not be possible without the aid of general anesthesia, the miraculous medical ability to turn off consciousness in a reversible and controllable way.

Researchers are using this powerful tool to better understand how the brain reconstitutes consciousness and cognition after disruptions caused by sleep, medical procedures requiring anesthesia, and neurological dysfunctions such as coma.

In a new study published in the journal eLife, a team led by anesthesiologists George Mashour, M.D., Ph.D. of University of Michigan Medical School, Michigan Medicine, Max Kelz, M.D., Ph.D. of the University of Pennsylvania Medical School, and Michael Avidan, MBBCh of the Washington University School of Medicine used the anesthetics propofol and isoflurane in humans to study the patterns of reemerging consciousness and cognitive function after anesthesia.

In the study, 30 healthy adults were anesthetized for three hours. Their brain activity was measured with EEG and their sleep-wake activity was measured before and after the experiment.

Each participant was given cognitive tests–designed to measure reaction speed, memory, and other functions–before receiving anesthesia, right after the return of consciousness, and then every 30 minutes thereafter.

The study team sought to answer several fundamental questions: Just how does the brain wake up after profound unconsciousness–all at once or do some areas and functions come back online first? If so, which?

“How the brain recovers from states of unconsciousness is important clinically but also gives us insight into the neural basis of consciousness itself,” says Mashour.

After the anesthetic was discontinued and participants regained consciousness, cognitive testing began. A second control group of study participants, who did not receive general anesthesia and stayed awake, also completed tests over the same time period.

Analyzing EEG and test performance, the researchers found that recovery of consciousness and cognition is a process that unfolds over time, not all at once. To the investigators’ surprise, one of the brain functions that came online first was abstract problem solving, controlled by the prefrontal cortex, whereas other functions such as reaction time and attention took longer to recover.

“Although initially surprising, it makes sense in evolutionary terms that higher cognition needs to recover early. If, for example, someone was waking up to a threat, structures like the prefrontal cortex would be important for categorizing the situation and generating an action plan,” says Kelz.

The EEG readings revealed that the frontal regions of the brain were especially active around the time of recovery. Importantly, within three hours of being deeply anesthetized for a prolonged period of time, participants were able to recover cognitive function to approximately the same level as the group that stayed awake during that time.

Furthermore, their sleep schedule in the days after the experiment did not appear to be affected.

“This suggests that the healthy human brain is resilient, even with a prolonged exposure to deep anesthesia. Clinically, this implies that some of the disorders of cognition that we often see for days or even weeks during recovery from anesthesia and surgery–such as delirium–might be attributable to factors other than lingering effects of anesthetic drugs on the brain,” says Avidan.

---

Modern anesthesia enabled increasingly complicated surgical and diagnostic procedures to be performed safely on patients, and has significantly advanced human medicine. For years after its advent, it was believed that general anesthetics (GAs) exert reversible, temporary effect on the central nervous system, which would return to its pristine state once the anesthetic exposure is ceased.

The long-lasting effects including cellular signaling changes and their impact after anesthetic exposure are enormous [1]. These effects can be desirable or undesirable. Indeed, anesthetics received during surgery were shown to be associated with brain dysfunction in young and elderly [2, 3].

In years to come, a large body of pre-clinical studies, and accumulating clinical evidences has steadily strengthened the belief that anesthetics may produce morphological changes and long-term functional impairment in brains at the extremes of age. Amidst the growing evidences linking GAs to neurocognitive impairment, the United States Food and Drug Administration issued a precautionary communication on GA use in patients aged three years and under [4], accentuating GA-related public health concerns. In this review, we attempt to provide a comprehensive discussion on the unwanted effects of general anesthetics on the central nervous system (CNS), integrating pre-clinical findings with clinical evidences.

Mechanism studies revealed that GAs act through various receptor proteins to modulate neuronal activities, to exert their amnesic, analgesic, sedative and immobilizing effects. The most recognized receptor targets include GABAA receptor (propofol, etomidate, isoflurane, sevoflurane), NMDA receptor (nitrous oxide, xenon, ketamine), glycine receptor and two-pore potassium channel [5, 6].

Such inhibitory and activating receptors are abundant throughout the mammalian brain, and may mediate unwanted, off-target effect of GAs to precipitate long-term cognitive dysfunction. In this regard, the extraordinary plasticity/connectivity and reduced compensating capacity of the developing and aging brains, respectively, may make them vulnerable to the ubiquitous, undesired actions of general anesthetics.

The developing brain
General anesthetics and neurotoxicity

Over the years, cellular and animal studies yielded substantial and convincing evidence on the cytotoxic and neurotoxic properties of general anesthetics. Since the pioneering study by Jevtovic-Todorovic et al., whereby 6 h exposure to a mixture of nitrous oxide, isoflurane and midazolam in postnatal day 7 rats induced long-term learning deficits [7], studies have demonstrated that routine GAs (isoflurane [8, 9], sevoflurane, propofol [10, 11], ketamine [12]) are capable of producing lasting cognitive, behavioral and memory deficiency in rodents when exposed in the early postnatal period.

Studies on non-human primates mirrored such findings, wherein early-life exposure to ketamine, sevoflurane or isoflurane led to persistent decline in cognitive, executive, memory and motivation-based tasks, and increased anxiety behaviors in the long term [13–15].

Based on the cumulative findings, the potency of GAs on neurobehavioral development is likely determined by the total length of exposure (a single lengthy vs. repeated brief exposures) and the developmental stage of the animal (first week postnatal). It was also reported that age of the neuron per se better predicts vulnerability to GAs than age of the organism, wherein juvenile neurons in adult animals are susceptible to the effect of GAs [16], to suggest neurocognitive toxicity of GAs even in adulthood.

The molecular mechanisms underlying GA’s lethality in developing neurons have been extensively explored. In vitro studies consistently reported the role of mitochondria and intrinsic (mitochondrial) apoptosis in GA-induced neurotoxicity. In neuronal culture and brain slice derived from immature rodents, isoflurane exposure significantly decreased anti-apoptotic BCL-2/pro-apoptotic Bax ratio, increased reactive oxygen species (ROS), and promoted cytochrome C release from mitochondria and caspase 3 cleavage [17–19].

Subsequent studies identified inositol 1,4,5-trisphosphate receptor (InsP3R) located on endoplasmic reticulum (ER) as a novel target of GA and an upstream signaling component of mitochondria. Under physiological conditions, activation of the InsP3R leads to Ca2+ release from ER lumen into the cytosol to initiate calcium-dependent signaling.

Isoflurane was shown to directly open InsP3R channels to induce excessive Ca2+ release from ER into cytosol and mitochondria, which further leads to mitochondrial calcium overload, ATP production failure, cytochrome C release and caspase activation [20–22]. In addition to targeting the mitochondria, recent data suggests that GA-induced cytosolic calcium buildup also impairs autophagosomal and autolysosome function to reduce cytoprotective autophagy, which would bias cell towards apoptosis [23] (Fig. 1).

Retrospective cohort studies found that multiple rounds of anesthetic exposure, and in young children under 2–4 years of age, were associated with learning difficulty and academic underachievement during childhood and adolescence [24, 25]. Single, brief anesthetics exposure, on the other hand, in pediatric patients younger than 3 years of age, was not found to be associated with neurocognitive or behavioral impairment [26]. However, one study reported that both single and multiple exposures to anesthesia were linked to language and abstract reasoning deficits [27].

The discrepancy is likely due to the selection bias inherent to retrospective study design, different assessment parameters, and/or age at assessment. Two prospective clinical studies examined the effect of single general anesthetic exposure at young age on future neurocognitive performance. The General Anesthesia compared to Spinal anesthesia (GAS) trial showed that GA is not associated with cognitive impairment compared to awake SA at 2 years of age [28].

The Pediatric Anesthesia Neurodevelopment Assessment (PANDA) trial also did not observe significant decline in cognitive, behavioral and memory capacity in GA-exposed subjects in comparison to their unexposed siblings, at 8–15 years of age [29]. Nevertheless, such findings cannot rule out the possibility that longer duration, repeated anesthetic exposure can harm the developing brain.

These studies are present with various confounding factors that warrant cautious interpretation of results. As anesthetics are rarely given alone, these studies rather assessed the association between surgery plus anesthetic exposure and cognitive/behavioral deficiency, instead the risks associated with anesthetics per se [30, 31].

In this regard, it would be difficult to dissect out the effect of surgery on neurocognitive development; moreover, children requiring surgery at young age are known to be different in many ways from those who do not, and such developmental differences may contribute to neurocognitive deficit attributed to surgery and/or anesthesia.

Furthermore, confounders such as hypotension, body temperature, and hypoxia during surgery are rarely described/controlled for in these studies, and could potentially alter the outcomes. In view of such, it would be very difficult to establish whether general anesthetics are causally linked to cognitive and behavioral deficiency, or conditions associated with such. Thus, large-scale observational studies and randomized trials with longer duration exposure of GAs and follow-up, more sensitive outcome measures, and stringent confounder control are required in the future, to provide more conclusive and informative data.

Neuroprotection in hypoxic-ischemic brain injury

Cerebral hypoxic brain injury contributes significantly to perinatal mortality and morbidity worldwide. It affects approximately 4 in 1000 births [32] and causes permanent neurological deficits in 25% of sufferers [33]. It is estimated that 4 million babies die in the neonatal period every year and birth asphyxia accounts for 23% of these deaths [34].

The lifelong consequences of perinatal hypoxic-ischemic encephalopathy to the affected infants, their family and the society necessitate the development of novel neuroprotective strategies. Hypoxic brain injury develops when oxygenation of the brain tissue is reduced, usually due to cardiac arrest or cerebrovascular incidents [35]. In the adult brain, this mostly occurs in the form of stroke.

In infants, the most common type of hypoxic brain injury is due to ischemia superimposed on hypoxia [33]. During or after birth, reduction in cerebral blood flow or further deoxygenation of the blood leads to the pathological asphyxia. The leading cause of hypoxic brain injury in the newborn is placental blood flow abruption and impaired gas exchange [36]. The brain injury is diffuse not focal, and affects the whole brain homogeneously [33].

During hypoxia/ischemia brain injury, energy depletion is due to the hypoxemia that switches cellular metabolism from aerobic to anaerobic. Anaerobic metabolism is insufficient to meet the cellular energy demands, which lead to depletion of stored ATP, creatinine phosphate and other forms of energy [37, 38]. Basic cellular proteins such as the Na+/K+-ATPase no longer function properly, leading to Na+ and Ca2+ influx, followed by cytotoxic edema and lysis [38, 39]. The brain tissue of the affected areas has a biphasic response to a hypoxic-ischemic injury [40, 41]. First, there is primary cell death, which includes the death of affected cells via necrosis during or shortly after the hypoxia, then secondary cell death via apoptosis 8–72 h after the hypoxia [42] or through autophagosomal or lysosomal death [43].

Glutamate neurotoxicity, or excitotoxicity, is the overstimulation of neuronal cells by glutamate that is released due to the depolarized membrane, and is a central feature to hypoxic-ischemic brain injury. Ischemic insult causes significant release of glutamate from excitatory nerve terminals, to promote water influx via the opening of surface channels such as the AMPA receptors and further influx of Ca2+ through the NMDA receptors on post-synaptic neuron membrane [44]. Accumulation of cytosolic Ca2+ leads to free radical production through formation of xanthenes and prostaglandins, cell membrane damage, phospholipase C activation, activation of endonucleases as well as apoptosis proteins, ultimately leading to cell death [45].

Much work has focused on the excitotoxicity-antagonizing effect of general anesthetic. Early studies reported that isoflurane directly interacts with the glutamatergic N-Methyl-D-aspartic acid (NMDA) receptor, whereby isoflurane may suppress NMDA or α-amino-3-hydroxyl-5-methyl-4-isoxazole-propionate (AMPA)-induced glutamate release, calcium uptake, mitochondrial membrane depolarization and excitatory neurotransmission [46–48].

Inhalational anesthetics sevoflurane and halothane were also shown to inhibit NMDA-induced excitotoxicity and calcium transient, however, the extent of neuroprotection appears to be lesser than that of isoflurane [48]. In this regard, excitotoxicity antagonism has been regarded as one of the mechanisms inhalational anesthetics act through to protect against ischemic brain injury. In addition, isoflurane interacts with and agonize inhibitory gamma-Aminobutyric acid-A (GABA) receptor that would inhibit depolarization and excitatory neurotransmission [49, 50] (Fig. ​(Fig.22).

Similar to that of volatile anesthetics isoflurane and nitrous oxide, xenon inhibits the plasma membrane Ca2+ pump, which might be responsible for neuronal Ca2+ concentration increase and altered excitability [51]. In 1998, it was shown that xenon suppresses nociceptive responsiveness through inhibition of NMDA receptors [52].

Xenon is different from all other volatile anesthetic agents, as it exerts no action on GABAA receptors [53]. It was predicted through biochemical modeling that xenon binds at the glycine site of the NMDA receptor and causes potent non-competitive inhibition [54]. Compared with commonly used anesthetic agents, xenon-induced anesthesia is featured with greater circulatory stability, lower analgesic consumption, lower adrenergic levels and better perfusion of individual organs [55]. Furthermore, xenon’s anesthetic effect is 1.5 times greater than that of nitrous oxide [56]. Nowadays, xenon has been used in anesthesia for many different types of surgery [57, 58].

Following the discovery that xenon is capable of inhibiting NMDA receptors, it was naturally postulated that xenon can protect neuronal cell against injury, since NMDA receptor-mediated neurotoxicity plays a critical role in neuronal cells death. Ma et al. [59] demonstrated the neuroprotective effect of xenon through N-methyl (D, L)-aspartate-induced neurotoxicity. Later, Ma et al. [60] demonstrated that xenon preconditioning improved both morphology and neurological functional outcome after the hypoxia–ischemia insults.

The mechanism of xenon preconditioning may be due to increased synthesis of survival proteins such as Bcl-2.The effect of xenon-mediated organoprotection was investigated in combination with methods in general clinical practice. Ma et al. [61] demonstrated that combination of xenon and hypothermia caused a synergistic enhancement of their individual neuroprotective properties. In addition to preconditioning, there are also studies demonstrating the effectiveness of post-treatment of xenon in brain injury. Dingley et al. [62] showed that xenon administered after a hypoxic-ischemic insult in neonatal rat model conferred uniformly 80% neuroprotection, as assessed by neuropathology of the major areas of the brain.

The neuroprotective effect of xenon was further tested on large animals. Schmidt et al. [63] evaluated whether xenon provides a neuroprotective effect to attenuate brain injury after transient cerebral ischaemia due to cardiac arrest in pigs. The major findings were that during reperfusion, brain injury is significantly smaller with the xenon treatment than control. Faulkner et al. [64] compared the effect of hypothermia and xenon-augmented hypothermia on the brain after transient global hypoxia-ischemia in piglet, whereby combination with xenon further reduced levels of cell death and tissue damage.

In 2010, xenon exposure combined with hypothermia was conducted for the first time on a newborn baby suffering from hypoxia at birth. The treatment was demonstrated to be very effective in attenuating mild brain injury in the young [65].

Hypoxia-inducible factor-1 (HIF-1) is the central mediator of the cellular response to hypoxic environments, and represents a key mechanism that inhalational anesthetics act upon to provide neuroprotection against ischaemic brain injury [66]. HIF-1 is a transcription factor belonging to the basic helix-loop-helix–Per-Arnt-Sim (bHLH–PAS) family.

It is a heterodimer composed of α and β subunits; α subunit is continuously made and degraded in both normoxic and hypoxic conditions, whilst β subunit is insensitive to oxygen [67]. HIF-1 mutations are known to lead to neural tube defects, brain underdevelopment and decreased neuronal cell number [68]. HIF-1 can be activated during hypoxia due to accumulation of the α subunit through reduced degradation [67]. The degradation of the HIF-1α subunit is mediated through the Von Hippel–Lindau (VHL) tumor-suppressor protein. VHL interacts with Elongin C and ubiquitinates HIF-1α, targeting it for proteosomal degradation [67]. HIF-1 hydroxylation is promoted by propyl hydroxylase (PHD), which is oxygen dependant [69]. Growth factors such as insulin-like growth factor (IGF)-1 and FGF bind to their receptors and activate the PI3 kinase/Akt/mTOR pathway, initiating HIF-1α production [70].

The critical gene expression after HIF activation mediates cellular responses to the hypoxia, including enhanced cell survival, erythropoiesis and angiogenesis [68] (Fig. ​(Fig.3).3). In this regard, volatile anesthetic halothane was shown to inhibit hypoxia-induced activation of HIF-1 [71], whereas isoflurane and xenon were demonstrated to upregulate HIF-1α expression and activity to confer kidney protection against ischemic insult [72–75]. Similarly, neuroprotection by isoflurane and xenon in ischemic brain injury is accompanied by HIF-1a upregulation [76, 77], which likely owes to volatile anesthetics’ ability to activate PI3K/AKT and ERK1/2 phosphorylation pathways.

reference link: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6443620/

---

Original Research: Open access.
“Recovery of consciousness and cognition after general anesthesia in humans” by George A Mashour, Ben JA Palanca, Mathias Basner, Duan Li, Wei Wang, Stefanie Blain-Moraes, Nan Lin, Kaitlyn Maier, Maxwell Muench, Vijay Tarnal, Giancarlo Vanini, E Andrew Ochroch, Rosemary Hogg, Marlon Schwartz, Hannah Maybrier, Randall Hardie, Ellen Janke, Goodarz Golmirzaie, Paul Picton, Andrew R McKinstry-Wu, Michael S Avidan, Max B Kelz. eLife