Abstract
General anaesthesia (GA) is a reversible drug-induced state of altered arousal required for more than 60,000 surgical procedures each day in the USA alone, making it one of the most common manipulations of the brain and central nervous system [1]. It comprises several specific behavioural and physiological end-points – unconsciousness, amnesia, analgesia and akinesia – with concomitant stability of the autonomic, cardiovascular, respiratory and thermoregulatory systems [2, 3].
The Hypnotic Effect
General anaesthesia (GA) is a reversible drug-induced state of altered arousal required for more than 60,000 surgical procedures each day in the USA alone, making it one of the most common manipulations of the brain and central nervous system [1]. It comprises several specific behavioural and physiological end-points – unconsciousness, amnesia, analgesia and akinesia – with concomitant stability of the autonomic, cardiovascular, respiratory and thermoregulatory systems [2, 3].
Many years after the first demonstration of anaesthesia in 1846 by William T. Morton in the ‘Ether Dome’ of the Massachusetts General Hospital in Boston, the mechanisms by which general anaesthetics render an individual unconscious remain incompletely understood [4].
For many decades, only one ‘general’ anaesthetic drug with a low therapeutic index [5] was used to induce immobility, analgesia, amnesia and unconsciousness. Nowadays, drugs with higher therapeutic indices and of different drug classes are available. Some combinations display synergism (Fig. 7.1, see also Chapter 3) which may be clinically useful by allowing us to use smaller doses of the individual drug (potentially decreasing side effects).Too much synergy may carry its own risk because there is synergy for side effects as well (e.g. ventilatory depression when benzodiazepines are combined with opioids) [6].
Fig. 7.1 Interaction grid summarizing drug interactions in humans and animals for hypnosis and immobility. Taken from [6], where detailed explanation of these interactions can be sought.
This chapter focuses on the hypnotic component of anaesthesia during induction and maintenance (unconsciousness and amnesia) and during emergence (i.e. regaining consciousness). Unconsciousness and amnesia are defined in Table 7.1.
Table 7.1. Differences between amnesia and unconsciousness. Adapted from [13].
| Hypnotic Component | Description |
|---|---|
| Amnesia | Partial or complete loss of memory. Memory formation occurs at various sites in the brain, including the hippocampus, amygdala, prefrontal cortex and other cortical sensory and motor areas. Patients should be unable to remember what happened during a surgical procedure. In this context, the possibility of both implicit and explicit memory formation is relevant. Explicit memory refers to information that is consciously perceived and retained, and can subsequently be reported. Patients who accidentally awaken from anaesthesia during surgery frequently suffer from an unpleasant experience that is recalled for a long time. Implicit memory refers to information that is unconsciously perceived, and that cannot be reported, but the retained information influences the patient’s behaviour without recollection. Anaesthetic-induced amnesia usually affects recall of events that occurred after the onset of anaesthesia (anterograde amnesia). |
| Unconsciousness | Conscious perception and/or response to environmental stimuli is impaired. Loss of responsiveness is not the same as loss of consciousness, even though sometimes they are used equivalently in the anaesthesia literature. For example, certain anaesthetics may produce behavioural unresponsiveness but not complete unconsciousness [11]. Responsiveness is usually used as a surrogate for consciousness, but it is still an indirect (and incomplete) measure of consciousness [12]. Furthermore, unconsciousness is a dynamic state. Its threshold should be continuously adjusted to the environment (e.g. noxious stimulation). For example, a patient receiving an infusion of propofol titrated to a surrogate measure of unconsciousness is not equivalent to the situation of surgical anaesthesia, because a noxious stimulus such as a scalpel cutting through skin could easily reverse this unconscious state [4]. |
Except when provided solely and exclusively by inhaled agents, the different components of GA are often provided by a mixture of hypnotic drugs and opioids (see Fig. 7.1). While specific drug classes target specific clinical end-points, i.e. hypnotics for hypnosis and ‘analgesics’ (opioids) for ‘analgesia’, opioids also have sedative effects and propofol also has analgesic properties [7]. The sedative effects of propofol result mainly from cortical disruption (discussed in more detail later), whereas those of opioids result from effects on acetylcholine, adenosine and dopamine receptors [3]. With regard to analgesia, propofol decreases the perception of pain related to a decreased state of consciousness, while opioids have an antinociceptive mechanism through a pathway-specific reduction of stimulus responses (modulation of afferent noxious stimulation as well) [7].
Several drugs are known to produce hypnosis: γ–aminobutyric acid type A (GABA-A) receptor agonists (propofol, etomidate, methohexital, thiopental, midazolam and diazepam), N-methyl-d-aspartate (NMDA) receptor antagonists (ketamine), α2-adrenoceptor agonists (dexmedetomidine, clonidine), μ-opioid receptor agonists (morphine, fentanyl, sufentanil, alfentanil and remifentanil) and dopamine receptor antagonists (droperidol and metoclopramide). Inhaled anaesthetics (halothane, enflurane, isoflurane, sevoflurane, desflurane, nitrous oxide [N2O] and xenon [Xe]) also induce hypnosis, the mechanisms of which remain enigmatic [6]. Considering this extensive list, it is fascinating to see that so many different drug classes and with such a diverse molecular structure can produce the same effect. Biologically, this phenomenon can be described as degeneracy, which is the ability of structurally different elements to perform the same function or yield the same output. Unlike redundancy, which occurs when the same function is performed by identical elements, degeneracy, which involves structurally different elements, may yield the same or different functions depending on the context in which it is expressed. It is a prominent property of gene networks, neural networks and evolution itself [8].
The mechanisms leading to anaesthetic-induced unconsciousness are complex and constitute a fundamental, unresolved question in both anaesthesiology and neuroscience [4]. Much evidence suggests that both consciousness and anaesthetic-induced unconsciousness depend on higher-order processes of the brain. Some models suggest an impaired capacity of the brain to integrate information across specialized subsystems as a common denominator. Several studies that investigated the effects of GA on the functional connectivity (FC) in the resting brain [9] illustrate that connectivity networks (cortical and subcortical) change, but the precise mechanisms remain controversial. One hypothesis is the ‘bottom-up’ paradigm, which argues that anaesthetics suppress consciousness by modulating sleep-wake nuclei and neural circuits in the brainstem and diencephalon that have evolved to control arousal states. The other is the ‘top-down’ paradigm, which argues that anaesthetics suppress consciousness by modulating the cortical and thalamocortical circuits involved in the integration of neural information [9, 10].
Targeting Optimal Delivery of Hypnotic Drugs
As described in Chapter 2, pharmacokinetic and pharmacodynamic (PK/PD) models allow clinicians to rationally dose and individually titrate anaesthetic drugs [14]. Parameters of several PK/PD models for several drugs are summarized in Table 7.2. The following section considers the clinically relevant parameters for the different anaesthetic drugs.
Table 7.2. Summary of the main anaesthetic drug models presented in this chapter. Model parameters are expressed as mean population values. PK: Pharmacokinetic; PD: Pharmacodynamic; CI: Confidence Interval; RSE: Relative Standard Error; SE: Standard Error; LBM: Lean Body Mass; TTPE: Time to Peak Effect.
| Anaesthetic Drug | Reference | Type of Model | Measure of Effect | Model Parameters |
|---|---|---|---|---|
| Ketamine | Dahan et al, 2011 [27] | PK/PD | Pain Scores | V1 (L) 53.26 (95% CI 38.71–69.59)V2 (L) 507.28 (95% CI 362.51–717.23)CL1 (L/per 70 kg) 83.34(95% CI 72.36–95.39)CL2 (L/per 70 kg) 118.32 (95% CI 85.62–162.84)EC50 (ng/mL) 10.5 (95% CI 4.37–21.2)γ 1.89 (95% CI 0.79–3.84)t½k (days) 10.9 (95% CI 5.25–20.50) |
| Ketamine | Schüttler et al, 1987 [25] | PD | Median Frequency of the EEG | Emax (Hz) 7.6±1.7EC50 (µg/mL) 2.0±0.5γ 3.8 ± 1.5 |
| Dexmedetomidine | Colin et al, 2017 [29] | PD | Bispectral Index of the EEG | keo (min−1) 0.12 (3.80 RSE %)EC50 (ng/mL) 2.63 (15.9 RSE %)ΔC50 1.71 (18.3 RSE %)Kin (min−1) 0.130 (24.6 RSE %) |
| Etomidate | Kaneda, Yamashita, Woo and Han, 2011 [31] | PK/PD | Bispectral Index of the EEG (also for the OAA/S) | V1 (L) 4.45 (7.4 SE %)V2 (L) 74.9 (41.7 SE %)CL1 (L/min) 0.63 (88.0 SE %)CL2 (L/min) 3.16 (21.4 SE %)keo (/min) 0.477 (9.9 SE %)t1/2keo (min) 1.55EC50 (µg/mL) 0.526 (14.9 SE %) |
| Propofol | Marsh et al, 1991 [15] | PK | V1 (L/kg) 0.228V2 (L/kg) 0.463V3 (L/kg) 2.893k10 (/min) 0.119k12 (/min) 0.112k13 (/min) 0.042k21 (/min) 0.055k31 (/min) 0.0033 | |
| Propofol | Schnider et al, 1998 [15] | PK/PD | EEG | V1 (L) 4.27V2 (L) 18.9–0.391x(age – 53)V3 (L) 238k10 (/min) 0.443+0.0107x(weight − 77)-0.0159x(LBM-59)+0.0062x(height − 177)k12 (/min) 0.302–0.0056x(age − 53)k13 (/min) 0.196k21 (/min) [1.29–0.024x(age − 53)]/[18.9–0.391x(age-53)]k31 (/min) 0.0035keo (/min) 0.456TTPE (min) 1.69 |
| Sevoflurane | Yasuda et al, 1991 [18] | PK | V1(L) 2.10±0.62k10 (/min) 1.78±0.32k12 (/min) 0.709±0.145k13 (/min) 0.223±0.035k14 (/min) 0.125± 0.056k15 (/min) 0.0310±0.0196k21 (/min) 0.194±0.092k31 (/min) 0.0231± 0.0198k41 (/min) 0.00313±0.00180k51 (/min) 0.000502±0.000117k20 (/min) 0.0094±0.0171 | |
| Sevoflurane | Olofsen and Dahan, 1999 [20] | PD | Bispectral Index of the EEG (also for Spectral Edge Frequency) | t1/2keo (min) 3.5±2.0Emax 94.5±3.1EC50 (%) 1.14±0.31γ 4.5±3.5 |
| Isoflurane | Yasuda et al, 1991 [18] | PK | V1(L) 2.31±0.71k10 (/min) 1.64±0.33k12 (/min) 1.26±0.204k13 (/min) 0.402±0.055k14 (/min) 0.243±0.072k15 (/min) 0.0646±0.0414k21 (/min) 0.210±0.082k31 (/min) 0.0230±0.0156k41 (/min) 0.00304±0.00169k51 (/min) 0.000500±0.000119k20 (/min) 0.0093±0.0137 | |
| Isoflurane | Olofsen and Dahan, 1999 [20] | PD | Bispectral Index of the EEG (also for Spectral Edge Frequency) | t1/2keo (min) 3.2±0.7Emax 95.7±2.2EC50 (%) 0.60±0.11γ 5.8±4.6 |
| Midazolam | Koopmans et al, 1988 [33] | PD | α – activity of the EEG | EC50 (ng/mL) 42.0 – 48.1γ 3.7 ± 1.8 |
Table 7.3. Components needed to define depth of anaesthesia. Reproduced with permission from [7].
| Afferent stimulus |
| Efferent response |
| Concentrations of analgesic components |
| Concentrations of hypnotic components |
| Concentrations of other relevant drugs (e.g. beta blockers, muscle relaxants, local anaesthetics) |
| Interaction surface relating the drug concentrations to the probability of the given response to the given stimulus |
Propofol
Several PK/PD models exist for propofol. Some have been integrated in commercially available target controlled infusion (TCI) systems, which in and by itself attests to the usefulness of these PK/PD models for propofol: TCI systems facilitate clinical practice and maintenance of intravenous anaesthesia [15].
Commercial TCI devices incorporate the Marsh and the Schnider propofol PK/PD models [15]. For effect-site targeting, the Schnider model should be used, for plasma targeting, the Marsh model.
The use of the lean body mass (LBM) equation in the Schnider model limits its use in obese patients. Different keo values have been proposed for effect-site targeting, but there is little conclusive evidence to demonstrate the superiority of any particular model or method. Clinicians should use the model and methods with which they are most familiar [15].
Effect-site targeting better maintained drug effect (evaluated with the bispectral index (BIS) as a measure of hypnotic effect) of propofol over time [16]. Out of three compartmental models (Marsh, Schnider, Schütller) and a physiologically based recirculatory PK model, the Schnider model (although not perfect) best predicted propofol plasma concentration with different infusion schemes (including TCI) [17] and is to be recommended, for TCI and advisory drug displays.
Sevoflurane and Isoflurane
Even though the PK of inhalational anaesthetics are most often described with a physiological model, empirical compartmental modelling has been used by Yasuda et al to describe sevoflurane and isoflurane PK [18]. The volume of the central compartment of sevoflurane (2.10±0.62 L) was smaller than that of isoflurane (2.31±0.71 L). The elimination rate constant from the central compartment was greater for sevoflurane than for isoflurane. The pulmonary elimination clearances and the rate constants from the peripheral compartments to the central compartment did not differ between the two drugs. The mammillary time constants were identical for all compartments except for the lung (smaller for sevoflurane). Estimated blood flow to the fat group was larger for sevoflurane. The recovery parameters for sevoflurane and isoflurane did not differ [18].
As part of the development of a model-based closed-loop control system, Gentilini et al built a PK/PD model for isoflurane (combined with alfentanil) and the BIS as measure of hypnotic effect [19]. Yasuda’s PK model was combined with a sigmoidal Emax model, yielding an EC50 of 0.6 % and a keo=0.217/min (or t1/2keo = 3.19 min). Olofsen et al reported the Emax models describing the relationship between isoflurane and sevoflurane effect-site concentrations and the BIS and the spectral edge frequency (SEF) of the EEG [20] to only differ with regard to their EC50 values: 0.6% for isoflurane versus 1.14% for sevoflurane using the BIS as end-point.
Others have confirmed keo values for sevoflurane and isoflurane to be the same [21], indicating that these two inhalational anaesthetic drugs have very similar PD profiles. McKay’s PK/PD model parameters describing the sevoflurane response and state entropy relationship ( t1/2keo = 2.65 mint1/2keo=2.65min and EC50= 1.7%) are similar to those reported by Olofsen above [20, 22]. The PD interaction between propofol and sevoflurane with BIS as the clinical end-point is additive, information that may be helpful to provide a smooth transition between the two anaesthesia techniques and to be used in advisory drug display systems [23].
Ketamine
Ketamine is an unique pharmacological agent with a turbulent history [24]. The racemic mixture R, S(±)-ketamine and its enantiomers S(+)-ketamine and R(–)-ketamine have different effects on the median spectral frequency of the EEG in humans [25]. The maximal decrease (mean ± SD) of the median frequency (Emax) for R,S(±), R(–) and S(+) ketamine were 4.4±0.5, 7.6±1.7 and 8.3±1.9 Hz, respectively, and the ketamine serum concentration that caused one-half of the maximal median frequency decrease (IC50) 2.0±0.5, 1.8±0.5 and 0.8±0.4 g/mL, respectively. In sheep, the time to peak effect or t1/2keot1/2keo of ketamine is identical to that of propofol [26]. A population PK/PD model for S(+)-ketamine-induced pain relief in humans has been described [27].
Dexmedetomidine
Dexmedetomidine is a selective and potent α2-adrenoceptor agonist with anxiolytic, sedative and analgesic properties. Existing PK/PD models have focused on healthy volunteers and ICU patients[28]. A PK/PD model for haemodynamic effects and for arousal and (BIS-guided) sedation in volunteers has recently been described [29, 30]. The models still have to be tested in a clinical environment. More studies are bound to have been published by the time this book goes to press.
Etomidate
Etomidate, a carboxylated imidazole ester, is an intravenous anaesthetic for which a PK/PD model with the BIS and the Alertness and Sedation (OAA/S) scale as a measure of effect has been described in healthy volunteers, but for which prospective testing in patients is still lacking [31]. In 20 volunteers receiving 5 mg/min etomidate eyelash reflex was lost after a mean etomidate dose of 24 mg. A two-compartment PK model described the concentration data well, no inter-individual variability was incorporated, the t1/2keot1/2keo was 1.55 min, and the Emax and EC50 of the sigmoidal Emax model were 67 and 0.526 µg/mL, respectively.
Propofol is faster than etomidate to induce loss of eyelash reflex and to lower BIS to 60, and also has a shorter intubating time [32].
Measuring the Effect
Prior to the induction of anaesthesia, consciousness of a subject can be easily assessed by a traditional clinical neurological exploration that includes verbal response, response to commands and brainstem reflexes. However, after induction of anaesthesia, when unconsciousness has been induced, one is forced to rely on surrogates to assess the hypnotic component of anaesthesia, even reflexes can no longer be used if neuromuscular blocking agents are used (the exception being pupillary reflexes). After initial loss of consciousness, unconsciousness has to be maintained and monitored in such a manner that both excessive and insufficient hypnotic effects are avoided. This is usually referred to as assessing depth of anaesthesia although it belongs directly to the hypnotic component of anaesthesia. The full assessment of adequacy of anaesthesia should involve a postoperative interview to exclude explicit recall.
The hypnotic state during anaesthesia can be measured to guide hypnotic drug targets and dosing by several methods and to several end-points, each of them useful in different phases of the procedure. These will be explained in more detail in this section.
Loss of consciousness
Unconsciousness is attained when the required amount of hypnotic drug reaches the presumably multiple required sites in the cerebral cortex, brainstem and thalamus that alter neurotransmission and induce hypnosis [33]. As the hypnotic drug concentration increases at these sites, individuals transition from wakefulness to unconsciousness. This transition process is a continuum that can be clinically observed as an all-or-none effect [34]. By careful assessment of the different stages it is possible to identify surrogates of the hypnotic effect.
Clinical End-points
The clinical end-points usually used as surrogates reflect both cortical, subcortical and/or spinal functions or indirectly measure their activity.
Cortical
In the continuum to unconsciousness we can determine the moment of loss of behavioural responsiveness (LOBR) to stimulation, when different strengths and types of stimuli cease to cause a response in the patient. This is mediated at the level of the cortex and happens because hypnotics render patients unwilling or unable to interact with their external environment [35]. Classifying the adequacy of the hypnotic effect based on cortical function is also possible using the Observers’ Assessment of Anaesthesia and Sedation score (OAAS/S). OAAS/S is a score that ranges from five (alert) to zero (unconsciousness) and includes end-points such as unresponsiveness to verbal commands and loss of response to shaking and shouting.
Subcortical
Hypnotics can have a profound impact on most actions of the nerves originating in the brainstem, actions that disappear during induction and reappear with recovery of consciousness. This supports the use of the brainstem reflexes as clinical end-points [36], namely the eyelash reflex, corneal reflex, oculocephalic reflex, gagging reflex and the pupillary light reflex [37]. These are non-specific indicators of impaired brainstem function due to the actions of the hypnotic agent on the oculomotor, trochlear, abducens, trigeminal, facial and glossopharyngeal nuclei in the midbrain, pons and medulla [34, 36].
The corneal reflex, for example, is elicited by mechanical stimulation of the cornea, resulting in the closure of the eyelid and requires the integrity of all these structures: (1) afferent nerves from the free nerve endings in the cornea; (2) the ocular division of the trigeminal nerve; (3) ganglion, root and spinal trigeminal tracts; (4) the spinal trigeminal nucleus in the pons. Then, some of the axons synapse into the (5) reticular formation interneurons, which send their axons bilaterally to (6) facial motor neurons in the facial nucleus (resulting in a consensual reflex), and subsequently through the (7) facial nerve, to the (8) orbicularis oculi, finally closing the eyelid.
Even though the assessment of different brainstem reflexes can provide useful insight into the actions of hypnotics in the brain and the integrity of the different structures, most anaesthesiologists use only the eyelash reflex, and even then, not consistently.
Spinal
Loss of muscular tone is also a possible clinical indicator to assess the hypnotic effect. It is usually tested by detecting the moment a patient drops an object held by the hand, a method known as ‘syringe drop’ because a 50 mL filled syringe was used in clinical studies. The rapid atonia that occurs after a bolus administration of most hypnotic agents can be due to their actions along the motor pathway and brainstem, but is most likely due to the actions of the drug in the spinal cord [34]. While muscle tone is also lost during REM sleep, brainstem reflexes are preserved.
Electroencephalography
There is increasing evidence that electroencephalography (EEG) can be used in the assessment of loss of consciousness [38, 39]. In a volunteer study [38] loss of consciousness coincided with an increase in low-frequency EEG power (< 1 Hz), the loss of spatially coherent occipital alpha oscillations (8–12 Hz), and the appearance of spatially coherent frontal alpha oscillations. These changes in the EEG reversed with recovery of consciousness. This study [38] also suggested that alpha amplitudes could predict the transition into and out of unconsciousness and profound unconsciousness.
In another volunteer study [39], the propofol concentration was gradually increased until the subjects were no longer able to interact with their external environment (= LOBR). After reaching LOBR, the amplitude of slow-wave activity (0.5 to 1.5 Hz) increased sharply, attaining saturation even though propofol concentrations continued to rise afterwards. Following discontinuation of the infusion, as drug levels dropped, the slow-wave power began to decrease and returned to baseline levels before recovery of the behavioural response. Therefore, it was proposed that slow-wave activity saturation could potentially be used as an individualized indicator of perception loss and be useful for monitoring the hypnotic component of depth of anaesthesia and studying altered states of consciousness.
Another approach to using EEG to identify the transition between consciousness and unconsciousness [40] combined data from other sources. This multimodal integration was entitled Anaesthesia Multimodal Index of Consciousness (AMIC), and comprised EEG, standard monitoring data (heart rate, arterial blood pressure, oxygen concentrations, end-expiratory carbon dioxide concentration, peak inspiratory pressure), individual patient data and drug information (MAC and/or plasma concentration). The AMIC not only identified the transition between consciousness and unconsciousness but was also able to separate out different levels of anaesthesia (from wakefulness to burst suppression).
Adequacy of Depth of Anaesthesia
Depth of anaesthesia assessment has been sought after since the beginning of anaesthesia itself. John Snow divided the effects of ether into five stages or degrees, which included end-points such as level of consciousness (awake to deep coma), muscle flaccidity and apnoea [41]. Guedel later refined these stages, introducing in 1937 the concept of different stages based on alterations in breathing, muscle tone, pupil diameter, lacrimation and eyelid reflex. Philip Woodbridge subsequently included the newer agents available in 1957 and presented an alternative way of looking at anaesthetic depth [42], which gained increasing interest with the widespread use of muscle relaxants. This spawned concerns regarding unintended intraoperative awareness, which is still a major concern nowadays. The incidence of awareness has been reported to range from 0.13% to 0.16% and can cause post-traumatic stress disorder in as many as 70% of these patients [43].
The impact of awareness on a patient’s well-being postoperatively has since long been recognized. Early methods proposed to assess depth of anaesthesia in an attempt to avoid it were based on clinical observations [42]. However, the assessment of adequacy of the depth of anaesthesia is very important not only in preventing intraoperative awareness, but also in tailoring drug delivery to each patient. This can reduce the amount of anaesthetic given, decrease the time to eye opening and extubation, and shorten the duration of postanaesthesia care unit stay [44].
We now know that depth of anaesthesia depends on other factors than those resulting from the action of hypnotics, thus it is important to recognize these other components, as Shafer and Stanski describe [7]:
1. Afferent stimulus
2. Efferent response
3. Concentrations of analgesic components
4. Concentrations of hypnotic components
5. Concentrations of other relevant drugs (e.g., beta blockers, muscle relaxants, local anaesthetics)
6. Interaction surface relating the drug concentrations to the probability of the given response to the given stimulus.
The assessment of the hypnotic effect during maintenance can therefore be combined with information gathered from other sources. For this there are several methods available.
Pharmacokinetic End-points
Minimum Alveolar Concentration –
One of the most widely used indicators to guide depth of anaesthesia, the minimum alveolar concentration (MAC), is a composite index of nociception and immobility as it represents the partial pressure of a volatile anaesthetic needed to prevent movement in 50% of the patients in response to a painful stimulus. This in fact shows the suppression of the withdrawal reflex, reflecting the volatile anaesthetic’s actions leading to immediate immobility via the spinal cord (Fig. 7.2).
