Local Anaesthetic Agents

Published on 27/02/2015 by admin

Filed under Anesthesiology

Last modified 22/04/2025

Print this page

rate 1 star rate 2 star rate 3 star rate 4 star rate 5 star
Your rating: none, Average: 0 (0 votes)

This article have been viewed 2037 times

Local Anaesthetic Agents

Local anaesthetics are analgesic drugs that suppress action potentials by blocking voltage-activated sodium ion (Na+) channels in excitable tissues. Examples include the anaesthetic amides (e.g. lidocaine, bupivacaine, ropivacaine) and esters (e.g. cocaine and procaine) (Table 4.1). Other drugs which can inhibit voltage-activated Na+ channels, such as diphenhydramine (a first-generation histamine H1 receptor antagonist) and amitriptyline (a tricyclic antidepressant) also have local anaesthetic properties. The blockade of voltage-activated Na+ channels accounts for both their analgesic effects, mediated through inhibition of action potentials in nociceptive neurones, and their systemic effects. The inhibition of action potentials in the heart contributes to local anaesthetic toxicity and also accounts for the antiarrhythmic actions of intravenous lidocaine (a class 1b antiarrhythmic). Unlike general anaesthetics (the pharmacology of which is described in Chs 1 and 2), local anaesthetics do not diminish consciousness when administered correctly.

Local anaesthetics block sensation at the site of administration by inhibiting action potentials in all nociceptive fibres and therefore do not discriminate between pain modalities, unlike other analgesic drugs, such as the anti-inflammatory agents and opioids. Opioid analgesics (morphine, fentanyl, hydrocodone, etc.) and other central analgesic drugs such as the α2-adrenergic agents (clonidine, dexmedetomidine) activate metabotropic (G protein-coupled) receptors within the membranes of specific neurones located within the pain pathway. A component of the actions of these drugs is centrally mediated (as described in Ch 5).

This chapter describes the pharmacology of local anaesthetics: their molecular mechanism of action, pharmacokinetics, systemic toxicity and recent developments which may improve their efficacy and safety.

MECHANISM OF ACTION OF LOCAL ANAESTHETICS

The primary target of local anaesthetics, the voltage- activated Na+ channel (VASC) is one of numerous membrane proteins which reside in the phospholipid bilayers encapsulating neurones (Fig. 4.1). VASCs provide selective permeability to Na+ when the cell becomes depolarized from the resting potential (approximately − 70 mV), which is maintained in quiescent neurones by the tonic activity of potassium ion (K+) channels. Local anaesthetics applied either topically to the skin or by infiltration inhibit action potentials in primary afferent nociceptive neurones, the pain-sensing neurones which transmit to the dorsal horn of the spinal cord (Fig. 4.2).

Pain transmission begins as a depolarization in the nerve ending of the primary afferent neurone initiated by the activation of cation channels. When the depolarization reaches the threshold for activation of VASCs (approximately − 45 mV), action potentials are generated, resulting in rapid depolarization to approximately + 20 mV (Fig. 4.3). Each action potential is brief (approximately 2 ms) because VASCs rapidly inactivate, leading to closure of their inactivation gates, and at the same time voltage-activated K+ channels activate, leading to an increase in the permeability of the cell membrane to K+. As a result, the membrane potential travels rapidly back towards the K+ equilibrium potential and this period is known as the after-hyperpolarization, a phenomenon which contributes to the refractory period during which it is unlikely that another action potential will be generated (Fig. 4.3).

Mechanism of Local Anaesthetic Inhibition of the Voltage-Activated Na+ Channel

Local anaesthetics inhibit VASC activity by gaining access to the open channel from the inside of the cell and binding to specific amino acids lining the channel lumen (Fig. 4.1). They bind preferentially to the open channel and are therefore said to be use-dependent (or open channel) blockers. First, the local anaesthetic must cross the cell membrane, a passage which requires lipid solubility. The molecule must then diffuse into the aqueous environment within the ion channel. Amide and ester local anaesthetics posses both lipophilic and hydrophilic properties and are described as amphipathic (Fig. 4.4). They exist in basic (uncharged) and cationic (charged) forms and the relative proportion of each (determined using the Henderson–Hasselbalch equation) is dependent upon the pH of the solution and the pKa of the local anaesthetic:

image

image

Local anaesthetics are weak bases and most have a pKa of approximately 8.5. Therefore there is approximately 10-fold more charged than uncharged molecule at physiological pH (~ 7.5):

image

image

An alkaline solution speeds the onset of analgesia by increasing the proportion of uncharged local anaesthetic on the outside of the nerve, resulting in more rapid access to the inside of the cell where the balance of isoforms re-establishes on the basis of the intracellular pH. By contrast, infected and inflamed tissue has a relatively low (acidic) pH leading to an increase in the proportion of the membrane-impermeant cationic local anaesthetic component and the requirement for higher doses to achieve analgesia.

THE VOLTAGE-ACTIVATED Na+ CHANNEL

Local anaesthetics gain access to their binding site within the inner lumen of the VASC when the activation gate opens in response to depolarization. The VASC is formed by a large protein (the α-subunit) consisting of 24 membrane-spanning segments arranged in four repetitive motifs (Fig. 4.1). The fourth segment of each motif is a voltage sensor, a series of positively charged amino acids (arginine and lysine residues) lying within the membrane. Depolarization causes electrostatic repulsion of the voltage sensors, providing the energy required to open the activation gate (Fig. 4.3). Na+ ions, selected by the filter formed by the four pore loops (between the 5th and 6th segments) lining the outer vestibule of the channel, are then free to pass down their concentration gradient into the cell, generating a depolarizing electrical current. However, Na+ current is inhibited by local anaesthetic bound within the inner vestibule of the channel. The inactivation gate, formed by intracellular components of the channel, closes rapidly following depolarization (Fig. 4.3) and local anaesthetics stabilize the inactivated state.

There are multiple subtypes of VASCs, named after the identity of their α-subunit (NaV1.1–NaV1.9) encoded by one of nine different genes (SCN1A–SCN5A, SCN8A–SCN11A) which are differentially expressed in tissues throughout the body and which have differing pharmacological and biophysical properties. This heterogeneity provides the potential (to date unmet) for selectively targeting VASCs in pain-sensing neurones. Nociceptive neurones predominantly express NaV1.7, NaV1.8 and/or NaV1.9 α-subunits. Mutations in the SCN9A gene, which encodes NaV1.7, are associated with several pain pathologies. Aspects of systemic toxicity relate to the ability of local anaesthetics to block VASCs outside the pain pathway. Cardiac VASCs are of the NaV1.5 subtype and local anaesthetics such as ropivacaine and levobupivacaine are thought to have less systemic toxicity due to their lower affinity for cardiac channels. Additional VASC heterogeneity is conferred by four genes which encode ancillary β-subunits.

PAIN FIBRES

Different peripheral nerve fibres have differing sensitivities to block by local anaesthetics and are classified as A, B and C according to their conduction velocities, A being the fastest conductors and C the slowest. Aδ and C fibres both conduct pain (Fig. 4.2). Other subtypes of A fibre supply skeletal muscles (α and γ) and conduct tactile sensation (β), while type B are preganglionic autonomic fibres. Aδ fibres are heavily myelinated and rapidly conduct acute stabbing pain. Myelination enables a remarkably high velocity of transmission (approximately 20 m s− 1) through a mechanism known as saltatory conduction. VASCs are segregated within the neuronal membrane of Aδ fibres at gaps in the myelin sheaths (nodes of Ranvier), enabling action potentials effectively to ‘jump’ from one node to the next. Aδ fibres are of small diameter and therefore have little ability to conduct changes in membrane potential once VASC activity has been inhibited. This makes them particularly sensitive to local anaesthetic block. Unlike Aδ fibres, C fibres are unmyelinated and their velocity of conduction from the skin to the spinal cord is relatively slow (approximately 1 m s− 1). Local anaesthetics effectively block the transmission of dull, aching pain mediated by C fibres. The fibre diameter is very small (approximately 1 μm) and therefore there is little passive conduction, making transmission reliant on the activity of VASCs. C fibres are activated by inflammatory mediators and therefore the pain resulting from their stimulation can also be treated by anti-inflammatory agents.

LOCAL ANAESTHETIC STRUCTURE

Local anaesthetics of the amide and ester classes share three structural moieties: an aromatic portion, an intermediate chain and an amine group (Fig. 4.4). The aromatic portion is lipophilic, and lipid solubility is further enhanced in local anaesthetics with lengthy intermediate chains. The amine group is a proton acceptor, providing the potential for both charged and uncharged isoforms (i.e. the source of the amphipathic nature of local anaesthetics). Amide and ester anaesthetics are so named because of their distinctive bonds within the intermediate chain. Examples of esters are cocaine, procaine, chloroprocaine and amethocaine; examples of amides are lidocaine, prilocaine, mepivacaine, etidocaine, bupivacaine, ropivacaine and levobupivacaine. A convenient mnemonic is that the names of esters contain one letter i while those of amides contain two letters i’s. The presence of either an amide or an ester bond dictates the pathway through which the local anaesthetic is metabolized and this has important implications regarding allergy potential and pharmacokinetic profile (described below). Since the introduction of cocaine into clinical practice by Koller in 1884, numerous local anaesthetics have been synthesized, beginning with procaine in 1905. Structural modifications influence pharmacokinetics. For example, replacement of the tertiary amine by a piperidine ring increased local anaesthetic lipid solubility and duration of action; addition of an ethyl group to lidocaine on the α-carbon of the amide link created etidocaine; and addition of a propyl group or butyl group to the amine end of mepivacaine resulted in [p]ropivacaine and bupivacaine respectively. Halogenation of the aromatic ring of procaine created chloroprocaine, an ester with faster hydrolysis and shorter duration of action.

PHARMACOLOGICAL PROPERTIES OF LOCAL ANAESTHETICS

A number of important factors influence the pharmacological profile of local anaesthetic drugs (Table 4.1). The pKa, molecular weight, lipid solubility, protein binding and vasoactivity influence speed of onset, potency and duration of action.

image pKa is the pH at which the ionized and non-ionized form of a compound is present in equal amounts. For basic drugs such as local anaesthetics, the greater the pKa, the greater the ionized fraction. As diffusion across the nerve sheath and nerve membrane requires non-ionized drug, a local anaesthetic with a low pKa has a fast onset of action while a high pKa causes a slow onset of action. For example, lidocaine (pKa 7.6) has a fast onset in comparison with bupivacaine (pKa 8.2), because, at pH 7.4, 35% of lidocaine exists in the non-ionized base form compared to only 20% of bupivacaine.

image Molecular weight influences the rate of transfer of drug across nerve membranes and through the dura mater. The lower the molecular weight the more rapid is the transfer.

image Lipid solubility, often expressed as the partition coefficient, influences potency. The partition coefficient is the ratio of aqueous and lipid concentrations when a local anaesthetic is introduced into a mixture of oil- and water-based solvents.

image Protein binding, including local anaesthetic attachment to protein components of the nerve membrane, increases the duration of action of a local anaesthetic. In plasma, amide anaesthetics bind predominantly to α-acid glycoprotein (AAG), a high-affinity limited capacity protein, and albumin, a low-affinity large capacity protein. The bioavailability of anaesthetic is determined by the availability of plasma proteins; the greater the AAG availability, the greater the binding of anaesthetic, and the lower the free plasma concentration. After surgery, trauma or malignancy, AAG concentrations increase significantly and protect patients receiving local anaesthetic epidural or perineural infusions from anaesthetic toxicity by curbing increases in the free fraction of local anaesthetics.

image Vasoactivity influences potency and duration of action. The vasoactivity of commonly used local anaesthetics is biphasic with dilatation occurring with anaesthetic concentrations ≥ 0.25% and vasoconstriction at concentrations < 0.25%. When measured by Laser Doppler flowmetry in the forearm, the vasoactive potencies occur in the order: lidocaine > bupivacaine > levobupivacaine > ropivacaine. Adrenaline at a dose of 1.25 μg provides significant vasoconstriction when administered with bupivacaine and levobupivacaine.

PHARMACOKINETICS

Absorption

The site, dose and rate of injection, and pharmacological properties, with or without addition of adrenaline, determine the absorption of a local anaesthetic drug from its site of injection. The maximum recommended clinical doses are shown in Table 4.2. The rank order of plasma concentration after injection at various sites is: intrapleural > intercostal > lumbar epidural > brachial plexus > sciatic > femoral, which reflects the vascular supply to these tissues. First-pass pulmonary metabolism limits the concentration of local anaesthetic which reaches the systemic circulation.

TABLE 4.2

Maximum Doses of Local Anaesthetics Administered as a Bolus

image

Adapted from McLeod GA, Butterworth JF, Wildsmith JAW (2008) Local anesthetic systemic toxicity. In: Cousins, Bridenbaugh, Horlecker, Carr (eds) Neural blockade. Lippincott, Williams & Wilkins, Ch 5 pp 114–132.

ENANTIOMER PHARMACOLOGY

Bupivacaine is a chiral molecule consisting of two structurally similar, non-superimposable, mirror images called enantiomers (Table 4.3). The nomenclature of enantiomers is based on the Cahn-Ingold-Prelog priority rules whereby the smallest atom is placed to the rear of the central atom about which the molecule rotates, and the sequence of the remaining three atoms is determined. For example, an increase in atomic mass in a clockwise direction is indicative of an S (sinistra) or laevo enantiomer whereas an increase in atomic mass in an anticlockwise direction is indicative of an R (rectus) or dextro enantiomer. Levobupivacaine is the S (–) or laevo enantiomer of bupivacaine whereas ropivacaine is a single enantiomer from the same series as bupivacaine, but with a propyl rather than a butyl group.

TABLE 4.3

Chiral Terminology

Chirality Spatial arrangement of atoms, non-superimposable on each other
Isomer Molecule with the same atomic composition but different stereochemical formulae and hence different physical or chemical properties
Stereoisomers Identical isomers which differ in the arrangement of their atoms in space
Enantiomer One of a pair of molecules which are mirror images of each other and non-superimposable
Racemate An equimolar mixture of a pair of enantiomers

INDIVIDUAL LOCAL ANAESTHETIC PHARMACOLOGY

LOCAL ANAESTHETIC TOXICITY

Scientific Investigations

Animal studies indicated that the dextro enantiomer of bupivacaine was considerably more toxic than the laevo enantiomer. In electrophysiology experiments, dextro-bupivacaine attenuated the maximum upstroke velocity of the action potential (Vmax) and lengthened action potential duration (APD), and in isolated small animal hearts it induced QRS widening, increasing both atrioventricular (AV) conduction time and arrhythmias.

In order to mimic inadvertent intravascular injection, lidocaine, bupivacaine, ropivacaine and levobupivacaine were infused into instrumented, conscious sheep in several studies designed to measure the cardiac and cerebral responses to systemic toxicity. The rank order of local anaesthetic dose needed to induce convulsions, arrhythmias, cardiac arrest and death was consistently: lidocaine > ropivacaine > levobupivacaine > bupivacaine. Importantly, arrhythmias induced by ropivacaine and levobupivacaine were more likely to revert to sinus rhythm, indicating that resuscitation may be easier in patients receiving these drugs.

Systemic Toxicity

Levobupivacaine and ropivacaine were introduced into clinical practice because both drugs showed equivalent potency for nerve block compared to bupivacaine but a greater therapeutic index from animal studies. However, systemic toxicity still remains a problem in clinical practice. Reasons for this include an increase in the practice of upper limb block, increased surgical use of local anaesthetics in high volumes for procedures such as tissue infiltration and tumescent anaesthesia, use of high-concentration compound local anaesthetic mixtures in the United States, inappropriate use of medical devices and administration of levobupivacaine and ropivacaine at doses greater than those recommended by the manufacturers.

Upper limb block is associated with a greater incidence of toxicity than lower limb or neuraxial block. For example, the incidence of convulsions secondary to epidural anaesthesia has been estimated at 1 in 8435, compared to 1 in 827 for axillary block, and 1 in 130 for supraclavicular and interscalene block. Tumescent anaesthesia for liposuction using doses of lidocaine > 50 mg kg− 1 has been associated with a mortality rate between 1 in 5000 to 10 000. Deaths have also been reported after application of 6–10% lidocaine and tetracaine compound local anaesthetic cream with cellophane wrapping to the legs before laser hair removal. In the UK, three patients died after unintentional connection of epidural infusions to intravenous lines, also between January 2005 and May 2006, the National Patient Safety Agency (NPSA) identified more than 346 incidents with epidural infusions which resulted in harm to patients. Most importantly, introduction of new ‘safer’ local anaesthetic drugs has encouraged some practitioners to administer higher doses of local anaesthetics than those recommended by drug manufacturers, narrowing the therapeutic window, and paradoxically increasing the risk of systemic toxicity.

Mechanisms of Systemic Toxicity

Direct injection into the vasculature (especially arterial injection in the head and neck) can lead to blindness, aphasia, hemiparesis, ventricular arrhythmias including fibrillation, convulsions, respiratory depression, coma or cardiac arrest. The most potent local anaesthetics have the highest tendency to cause systemic toxicity. For example, bupivacaine is more toxic than lidocaine. Cardiovascular effects are caused by blockade of cardiac VASCs and K+ channels. Levobupivacaine and ropivacaine are thought less likely to interact with cardiac VASCs. Convulsions may be caused by the blockade of GABAA receptors in the CNS and respond to positive modulators of GABAA receptor function (barbiturates, propofol and benzodiazepines). Involvement of mitochondria in local anaesthetic toxicity was proposed when a patient with carnitine deficiency showed marked sensitivity to a low dose of bupivacaine, suggesting that local anaesthetics interfere with mitochondrial energy functions. Animal experiments and many case reports have shown the benefits of Intralipid infusion during resuscitation following local anaesthetic-induced cardiac arrest.

MANAGEMENT OF SEVERE LOCAL ANAESTHETIC TOXICITY

Treatment of systemic toxicity is outlined in the guidelines produced by the Association of Anaesthetists of Great Britain and Ireland (AAGBI) (Table 4.4).

TABLE 4.4

Management of Severe Local Anaesthetic Toxicity: AAGBI Safety Guideline

image

Reproduced with permission from AAGBI Safety Guideline Management of severe local anaesthetic toxicity http://www.aagbi.org/publications/guidelines/docs/la_toxicity_ 2010.pdf.

Recognition of the prodromal symptoms, such as agitation and disruption of sensory perception, is essential in order to instigate full life support measures. Immediate management should follow the Airway, Breathing, Circulation (ABC) mnemonic of resuscitation as outlined by the UK Resuscitation Council. Seizures should be controlled with small doses of thiopental, propofol or midazolam, depending on what is at hand. Hypoxaemia, hypercapnia and acidosis should be avoided because they all suppress myocardial function. If cardiac arrest occurs, cardiopulmonary resuscitation (CPR) should be started using standard protocols, but it should be recognized that arrhythmias may be very refractory. If so, consideration should be given to using an intravenous bolus dose of 20% Intralipid 1.5 mL kg− 1 over 1 min followed by an infusion of 15 mL kg h− 1. If still unresponsive, a further two boluses may be given and the infusion rate doubled (Table 4.5). The role of adrenaline in lipid-based resuscitation from local anaesthetic-induced cardiac arrest is the subject of debate and is still to be resolved. Several case reports now allude to the success of Intralipid in both paediatric and adult resuscitation.

TABLE 4.5

Intralipid Doses

image

Do not exceed a maximum cumulative dose of 12 mL kg− 1.

Reproduced with permission from AAGBI Safety Guideline Management of severe local anaesthetic toxicity http://www.aagbi.org/publications/guidelines/docs/la_toxicity_ 2010.pdf.

Prevention of Severe Local Anaesthetic Toxicity

image Regional blocks should always be performed in an area equipped to deal with cardiorespiratory collapse, such as an anaesthetic room or block room within the theatre suite.

image The age, weight and infirmity of the patient should be taken into account, and doses adjusted accordingly.

image Syringes of local anaesthetics and perineural and epidural infusions should be labelled clearly. Use of premixed sterile solutions is encouraged.

image Gentle aspiration of the syringe should precede every injection, but anaesthetists should be aware that negative aspiration does not guarantee extravascular positioning of the needle tip – false negatives do occur.

image Both during and after drug administration, the anaesthetist must keep talking to the patient.

image An appropriate test dose should be given depending on the situation. For example, a test dose of 3 mL of ‘epidural’ bupivacaine 0.5% (15 mg) injected accidentally into the intrathecal space will provide a definitive outcome – spinal anaesthesia. In contrast, injection of 0.5 to 1 mL during a perineural block under ultrasound is usually sufficient to differentiate between intraneural and extraneural injection.

image Examples of alternative epidural test doses are lidocaine 2% or adrenaline 15 μg, but neither test is specific or sensitive

image Ultrasound allows visualization of the position of the needle or catheter, their relationship to other structures – both nerves and large blood vessels – and the spread of local anaesthetic solution, although no definitive evidence exists yet that its use reduces overall complication rates.

EMERGING LOCAL ANAESTHETIC APPROACHES

Imaging and Local Anaesthesia

Ultrasound has become an important 2-D technique for guiding regional anaesthesia. Direct visualization of peripheral nerves, blood vessels and muscle is now possible during needle insertion both in-plane and out-of-plane of the ultrasonic beam. Local anaesthetic is deposited precisely around nerves using the hydrolocation technique whereby 1-mL boluses of local anaesthetic are used to distend connective tissue. The spread of local anaesthetic on a standard ‘B-Mode’ image is relatively easy to see and is quite distinct from the intraneural swelling characteristic of direct intraneural injection in laboratory preparations. However, despite undoubted improvements in block efficacy, no evidence as yet exists to demonstrate that the incidences of inadvertent intraneural or intravascular injection have declined.

TRP Channels and Pain

There has been a great deal of recent interest in targeting ion channels of the transient receptor potential (TRP) family to produce analgesia. Noxious heat activates TRP channels of the TRPV1 subtype in primary afferent nociceptive neurones. TRPV1 channel activation triggers the influx of Na+ and Ca2 + ions, leading to depolarization and activation of VASCs. The resulting action potentials in pain fibres trigger the burning pain stimulus (Fig. 4.2). TRPV1 channels are also activated by the vanilloid capsaicin, a pungent substance extracted from chilli peppers. Other subtypes of TRP channel (TRPM8 and TRPA1) appear to be stimulated by noxious cold. There is interest in antagonists of TRP channels as potential analgesic drugs. Recent studies also demonstrate that TRPV1 channels can flux the positively charged quaternary lidocaine molecule QX-314. The drug is permanently charged and therefore cannot pass across the membrane of nociceptive neurones without an aqueous passage. Stimulation of TRPV1 channels by capsaicin provides a conduit for entry of QX-314 into pain fibres. Once inside neurones, the charged molecule binds to the local anaesthetic binding site within VASCs and inhibits action potentials. There is interest in this approach for targeting local anaesthesia to nociceptive neurones, thereby increasing the selectivity of block. Animal studies demonstrate that this method provides prolonged analgesia with much less motor block than is caused by equivalent analgesia using lidocaine.

FURTHER READING

AAGBI Safety Guideline. Management of severe local anaesthetic toxicity. http://www.aagbi.org/publications/guidelines/docs/la_toxicity_ 2010.pdf

Binshtok, A.M., Bean, B.P., Woolf, C.J. Inhibition of nociceptors by TRPV1-mediated entry of impermeant sodium channel blockers. Nature. 2007;449:607–610.

Catterall, W.A., Dib-Hajj, S., Meisler, M.H., Pietrobon, D. Inherited neuronal ion channelopathies: new windows on complex neurological diseases. J. Neurosci. 2008;28:11768–11777.

Chang, D.H., Ladd, L.A., Wilson, K.A., Gelgor, L., Mather, L.E. Tolerability of large-dose intravenous laevo-bupivacaine in sheep. Anesth. Analg. 2000;91:671–679.

McLeod, G.A., Burke, D. Levobupivacaine. Anaesthesia. 2001;56:331–341.

Rao, R.B., Ely, S.F., Hoffman, R.S. Deaths related to liposuction. N. Engl. J. Med. 1999;340:1471–1475.

Rosenblatt, M.A., Abel, M., Fischer, G.W., Itzkovich, C.J., Eisenkraft, J.B. Successful use of a 20% lipid emulsion to resuscitate a patient after a presumed bupivacaine-related cardiac arrest. Anesthesiology. 2006;105:217–218.

Strichartz, G.R., Sanchez, V., Arthur, G.R., Chafetz, R., Martin, D. Fundamental properties of local anesthetics. II. Measured octanol: buffer partition coefficients and pKa values of clinically used drugs. Anesth. Analg. 1990;71:158–170.

Weinberg, G., Ripper, R., Feinstein, D.L., Hoffman, W. Lipid emulsion infusion rescues dogs from bupivacaine-induced cardiac toxicity. Reg. Anesth. Pain Med. 2003;28:198–202.

Yanagidate, F., Strichartz, G.R. Local anesthetics. Handb. Exp. Pharmacol. 2007;177:95–127.