CHAPTER 17 Principles of General Anesthesia
The pioneering use of anesthetics is credited to two dentists: Horace Wells and his one-time pupil and partner, William T.G. Morton, who practiced dentistry in New England in the early 1800s. Their achievements were preceded by the contributions of many others and came at a time when still others were carrying out experiments that would lead them to compete for recognition as the discoverers of anesthesia.
The history of anesthesia is no doubt as old as humankind itself, for surely since the dawn of time people have sought ways to alleviate pain. Records spanning thousands of years make it clear that patients about to undergo painful procedures have sought recourse in prayer; magic; the intervention of witch doctors and medicine men; techniques such as compression of nerves and blood vessels; and various plant products such as opium, mandragora, and coca. Modern anesthesiology had its beginnings in the eighteenth and early nineteenth centuries. The development of physics and chemistry led to the discovery of elements and simple molecules, including numerous gases. Joseph Priestley, an English scientist, is credited with the discovery of carbon dioxide, oxygen, and, in 1772, nitrous oxide. Although he thought oxygen might have some medical use, Priestley was unaware of the anesthetic properties of nitrous oxide. In 1795, Humphry Davy, a 17-year-old surgeon’s assistant in England who later became a distinguished scientist himself, began experiments with nitrous oxide. He inhaled the gas and used it on one occasion to relieve the pain of his erupting third molar (although at this time nitrous oxide was still considered to be extremely poisonous). He noted in his published studies of nitrous oxide the giddiness, pleasurable sensations, relaxation of muscles, and diminution of pain that were produced by inhalation of the gas. In 1799, Davy constructed the first machine for the storage and inhalation of nitrous oxide.
The development of anesthesia was carried further by Michael Faraday, Davy’s student, who in 1818 noted the anesthetic properties of diethyl ether (then known as “sweet vitriol”), and by Henry Hills Hickman, an English surgeon who carried out painless surgery on laboratory animals with carbon dioxide gas as the anesthetic. In 1824, Hickman published a pamphlet, “A Letter on Suspended Animation,” in which he suggested that patients could be made unconscious before surgery.
In the United States in the early 1800s, there was scientific and popular interest in ether and nitrous oxide. Itinerant entertainers who called themselves professors went about delivering lectures on these substances and demonstrating their effects. One of the earliest of these demonstrations was conducted in 1824 by Joseph Dorfeuille, a museum director from Cincinnati, who gave nitrous oxide to a dozen spectators. “Laughing gas” parties and “ether frolics” became common among medical students; because of his experiences at such an ether party, William E. Clarke, one such medical student, administered ether from a towel to a young woman having a tooth extracted in Rochester, New York. This use of ether in 1842 is the first on record.
Crawford W. Long, a Georgia physician who had been trained at the University of Pennsylvania Medical School, had attended ether frolics while a student, and, later in 1842, he used ether when he removed two small tumors from the neck of James Venable, a friend who had previously experienced the effects of inhaling ether. Credit for the first use of ether in a nondental procedure belongs to Dr. Long. His anesthetic fee, also the first on record, was $2. Because Long wanted to include observations of the effects of ether in major surgical procedures, he did not publish reports of his pioneering use of ether until 1849, 3 years after the accounts of Morton’s use of ether had appeared. A letter from Long written in 1844 suggests that he was visited by a dentist and a surgeon from Boston, that the dentist was Morton or Wells, and that it was from Long that they learned the technique of administering ether during surgery.
On December 10, 1844, Horace Wells attended a demonstration, staged by Gardner Quincy Colton in Hartford, Connecticut, of the effects of “laughing gas.” One subject who volunteered to take the gas injured himself in the leg. Wells noticed that he was unaware of his injury and apparently had no pain until the effects of the gas wore off. The next day, Wells persuaded John Riggs, a prominent Hartford dentist, to remove one of his own teeth while under nitrous oxide anesthesia administered by “Professor” Colton. Wells claimed that he felt no more than a pinprick. Wells then obtained permission to demonstrate his technique before a class at the Harvard Medical School and administered nitrous oxide to a student, who proceeded to scream loudly while his tooth was being removed. The boy later said he had felt no pain. Discouraged by the apparent failure of his demonstration and by the hostile reception that followed, Wells became ill and was unable to practice dentistry on a regular basis. He nevertheless continued to administer nitrous oxide, with mixed success, for dental and medical operations. Wells also experimented with ether in 1845 and with chloroform when its anesthetic effect became known (in November 1847). Wells died in January 1848 when he became deranged by overexposure to chloroform and committed suicide while in jail for having accosted a prostitute. Nitrous oxide was abandoned after his death until 1863, when Colton reintroduced its use for dental extractions.
William T.G. Morton of Boston, a former student and partner of Wells, had begun to use ether topically for its local numbing effect on his dental patients. With the help of his chemistry professor at Harvard, Charles T. Jackson, Morton refined his technique and successfully administered anesthesia to a patient for the extraction of a molar tooth. Convinced of the importance of his discovery, he obtained an invitation to demonstrate his technique for John C. Warren, a surgeon at Massachusetts General Hospital. On October 16, 1846, Morton prepared a young patient for the surgical removal of a large mandibular tumor. Morton is credited with the discovery of anesthesia and the custom of saying, “Doctor, your patient is now ready.”
Morton was anxious to patent the substance he called “Letheon,” but several physicians from the Massachusetts General Hospital thought it unsuitable to patent a medical discovery and indicated that they would not continue to use it if its chemical nature remained a secret. Morton then offered to make known the nature of the substance and to serve as an anesthetist at various hospitals. He abandoned his medical studies and his dental practice and became the first professional anesthetist. In 1846, Holmes addressed a letter to Morton suggesting that the term anesthesia be given to the state produced by ether and that the agent itself be called an anesthetic.
After Morton’s demonstration in Boston, the use of anesthesia spread rapidly despite opposition from various groups, many of whom still believed that there was something spiritually ennobling about pain, particularly the pain of childbirth. In 1847, James Young Simpson first used ether in his obstetric practice and in the same year successfully delivered a child using chloroform. Later, when Queen Victoria delivered her seventh child while under chloroform anesthesia, most ecclesiastic opposition was stilled.
No new anesthetic agents were added until the 1920s and 1930s, when ethylene, cyclopropane, and divinyl ether were introduced. Since the early 1950s, a series of halogenated agents containing fluorine have been introduced clinically and have essentially replaced other inhalation agents except nitrous oxide.
Intravenous agents, mainly the thiobarbiturates (e.g., thiopental), became popular in the late 1930s. Other ultrashort-acting barbiturates were added to the list and were supplemented in the late 1960s by ketamine and the neuroleptanalgesic combination of droperidol-fentanyl. Additional newer intravenous anesthetics include etomidate, midazolam, and propofol.
Neuromuscular blocking drugs were added to the practice of anesthesia in 1942 with the introduction of curare to facilitate endotracheal intubation and relaxation of muscles for abdominal surgery. Opioid anesthesia, in which morphine and subsequently fentanyl and its congeners found use as principal agents in obtunding autonomic responses to surgical stimulation, originated with cardiac surgery in the late 1950s. Dexmedetomidine, a centrally acting α2-adrenergic receptor agonist related pharmacologically to clonidine, represents yet another approach to providing sedation and analgesia during operative procedures.
Finally, the use of nitrous oxide by dentists has exhibited a cyclic pattern of popularity every 25 to 30 years since Wells first used it. Nitrous oxide is currently enjoying an extended fifth cycle as a sedative agent. As an agent for general anesthesia, nitrous oxide is slowly losing popularity, however, for reasons described subsequently and in Chapter 18.
General anesthesia may be defined as “a drug-induced reversible depression of the CNS resulting in the loss of response to and perception of all external stimuli.”4 In practice, this simple definition is inadequate because it neglects the contributions of unconsciousness, amnesia, immobility, and autonomic stability to the anesthetic state and the fact that general anesthetics differ significantly in the effects they achieve.
A complete anesthetic is one that produces unconsciousness, unresponsiveness, amnesia, analgesia, and muscle relaxation by itself without eliciting undue homeostatic disturbances in the patient. An example of such a complete anesthetic is diethyl ether (known simply as ether). Although there are other complete anesthetics, the tendency in modern anesthesiology is to use a combination of drugs to take advantage of the best properties of each and to minimize unwanted side effects. Combining anesthetics from different drug classes allows for a reduction in the dose of each agent as the majority of such interactions are supra-additive in nature.
Among the agents that may be used preoperatively are the antimuscarinic drugs to minimize salivation, laryngospasm, and reflex bradycardia and various analgesics and CNS depressants to provide preoperative pain relief, sedation, and amnesia. Drugs that are used during the administration of general anesthesia in addition to the primary anesthetic may include nitrous oxide; intravenous opioids (which lessen the total required dose of anesthetic and increase analgesia); midazolam or another amnestic drug to prevent recall; drugs that paralyze skeletal muscle; antiemetics such as ondansetron to limit postoperative nausea and vomiting; and, if necessary, drugs that help maintain cardiovascular stability and renal function.
The primary goals of general anesthesia are to preserve the life of the patient, to provide the operator with an adequate surgical field, and to obtund pain. A general anesthetic ideally should (1) provide a smooth and rapid induction; (2) produce a state of unconsciousness or unresponsiveness; (3) produce a state of amnesia; (4) maintain essential physiologic functions while blocking reflexes that might lead to bronchospasm, salivation, and arrhythmias; (5) produce skeletal muscle relaxation, but preferably not of the respiratory muscles, through the blockade of various efferent impulses; (6) block the conscious perception of sensory stimuli so that there is adequate analgesia to perform the procedure; and (7) provide a smooth, rapid, and uneventful emergence and recovery with no long-lasting adverse effects.
The goals of anesthesia for general surgery also apply to dental surgery, but there are some important differences. Dental patients are generally outpatients; in most circumstances, particularly situations not involving extensive oral surgery, the procedures are not as traumatic as general surgical procedures, and it is neither necessary nor desirable to render the patient unconscious. Although general anesthesia is sometimes necessary, specific techniques have been developed for producing sedation in dental patients (see Chapter 48).
Since the introduction of general anesthetics, considerable efforts have been directed toward discovering the mechanism of action of these agents. Our incomplete knowledge of the structure and behavior of membrane constituents and subcellular organelles, neurotransmitters, and neurologic circuits, coupled with an imperfect understanding of behavioral states relevant to clinical anesthesia such as consciousness, sleep, pain, and anxiety, makes elucidation of what causes anesthesia extremely difficult. Information has been gleaned for each of the above-mentioned functions, however, and is summarized in the following sections.
Many investigators have sought to describe the action of the extremely diverse chemicals known to be general anesthetics by their ability to perturb the molecular structure and function of neurons. Most early anesthetic agents seemed to be indiscriminate in affecting biophysical properties of cellular and subcellular membranes, and for many years it was generally agreed that there were no specific receptors for general anesthetics (and therefore no direct antagonists) as there are for neurotransmitters. In this setting, a universal mechanism of action of general anesthesia based on the physicochemical properties of anesthetic agents was postulated. More recently, many actions of general anesthetics have been documented, and it is now believed that diverse molecular perturbations may result in unconsciousness and lack of response to external stimuli.
Various mechanistic theories of general anesthesia began to appear shortly after the landmark demonstration of ether-induced insensibility by Morton, but the first important observation was made independently by Meyer in 1899 and Overton in 1901, who emphasized the correspondence between the lipid solubility of an agent and its anesthetic potency (Figure 17-1). The Meyer-Overton correlation suggested that anesthesia begins when any chemical substance has attained a certain molar concentration in the hydrophobic phase of the cell membrane. When olive oil is used to represent the hydrophobic medium, this concentration is approximately 50 mmol/L. Experiments with different lipid media indicate that the best fit between solubility and anesthetic potency is obtained with lipids that are amphophilic (i.e., they have polar and nonpolar attributes) and can serve as hydrogen bond acceptors. These characteristics are descriptive of membrane phospholipids and cholesterol.
FIGURE 17-1 Linear correlation between anesthetic potency and lipid solubility. Potency is indicated by the minimum alveolar concentration (MAC) and lipid solubility by the olive oil/gas partition coefficient.
In 1954, Mullins, in his critical volume hypothesis, modified the original correlation to include consideration of the volume of the hydrophobic region occupied by the anesthetic agent. He reasoned that large anesthetic molecules would have greater effects on the membrane than would smaller molecules.
Numerous investigators since Mullins have sought to link the notion of a critical number or volume of anesthetic molecules with plasma membrane disturbances that could result in general anesthesia. Until the early 1980s, most attention was directed at the lipid bilayer of the plasma membrane, specifically the ability of anesthetics to cause membrane expansion, lipid fluidization, or lateral phase separation. With each of these effects, it was postulated that, as a result of the alteration in the lipid bilayer, the neuronal membrane becomes unable to facilitate the changes in protein configuration that are required for such essential steps in the transmission of nerve impulses as ion gating, synaptic transmitter release, and binding of the transmitter to the receptor.31
The membrane expansion theory was a natural outgrowth of the critical volume hypothesis. It holds that the absorption of anesthetic molecules by the lipid phase causes the membrane to expand, preventing important intrinsic membrane constituents from functioning properly. Measurements indicate that the expansion associated with general anesthesia is approximately 0.4%. Fluidization, or disordering, of lipids by anesthetic agents was noted in studies of lipid bilayers prepared with phospholipid and cholesterol to mimic cell membranes. Parallel shifts in measures of lipid fluidization and the activity of membrane-bound enzymes suggested that this perturbation of the normal lipid structure may result in functional changes sufficient to disrupt nerve transmission. The lateral phase separation theory was based on the idea that membrane lipids exist in two states: a high-volume, disordered sol state and a compact, ordered gel state.31 The ability of lipids to convert from the sol to the gel configuration, or to be compressed laterally within the membrane, was thought to accommodate conformational changes that need to occur for the opening of ion channels.
These lipid perturbation theories were supported by findings that hyperbaric pressures and certain convulsant drugs antagonize anesthesia, presumably by reversing membrane expansion or re-establishing order. It is now understood, however, that pressure or drug reversal of anesthesia arises from a physiologic antagonism of anesthetic action brought on by independent neurologic stimulation. Different anesthetics are affected differently by the same pressure, including chloral hydrate, whose anesthetic effect is immune to pressure reversal. Evidence has also mounted to cast doubt on membrane expansion or lipid perturbation per se as a cause of anesthesia. Direct measurements of the expansion of lipid bilayers and red blood cell membranes in response to anesthetic concentrations of ethanol and halothane yield values that are effectively insignificant, and other measurements have shown nonanesthetic long-chain alcohols to cause membrane expansion similar to that of inhalation anesthetics. Regarding fluidization or sol-gel transformations, changes equivalent to those associated with anesthesia can be attained by temperature elevations less than 1° C.32
Calculations based on the Meyer-Overton relationship argue in general against a significant effect of anesthetic drugs on membrane lipids. At concentrations sufficient to produce surgical anesthesia, there is only about one molecule of drug in the membrane for every 60 to 80 molecules of the much larger lipid constituents. Unless anesthetic molecules are distributed unevenly in the membrane (e.g., concentrated in lipids adjacent to ion channels) or the lipid phase serves as a barrier to the diffusion of anesthetic agents (i.e., limiting access of anesthetics to their effector site) or as a reservoir for them (i.e., retaining anesthetic molecules where they have direct access to their effector sites), it is unlikely that membrane lipids play a major role in the mechanism of anesthesia.
Membrane proteins constitute a second hydrophobic environment with which anesthetic molecules may interact.9 The idea that membrane proteins are the targets of anesthetic action is attractive for several reasons. First, it is consistent with the mode of action of most drugs that influence the CNS. Second, allosteric selection (described in Chapter 1) of a protein conformation by the binding of even a single small molecule can have pronounced effects on protein function. Third, it can best explain differences in action among the various anesthetics by assuming that these agents exert different effects on the same protein or influence different proteins altogether.
Although technical difficulties have long inhibited direct examination of anesthetic drug interactions with membrane proteins, the firefly enzyme luciferase provided a good initial model for study.7 Luciferase is a water-soluble protein that produces light when it cleaves its substrate, luciferin. For a wide variety of agents, anesthetic potency correlates directly with the ability to inhibit luciferin binding and prevent light emission. The binding site on the enzyme is amphophilic in nature and capable of accepting a hydrogen bond. Similarly, numerous potential sites exist for direct anesthetic interactions with proteins: hydrophobic regions within globular or folded polypeptides, between polypeptides joined in an oligomeric structure, and at the protein-lipid or protein-water interface.
The synthesis of a water-soluble four-stranded α-helical protein bundle similar in structure to peptides forming ligand-gated ion channels has allowed characterization of direct anesthetic-protein interactions.14 A hydrophobic cavity within the bundle binds halothane best when the cavity is lined with methionine and aromatic acid residues. Halothane binding stabilizes the protein in a conformation that putatively promotes anesthesia.
A close correspondence between the anesthetic potencies of stereoisomers of halothane and isoflurane and their ability to perturb ion channel function provides strong evidence that these membrane proteins are the immediate targets for general anesthetic action. It is now firmly established that certain classes of general anesthetics inhibit or activate specific ligand-gated ion channels in clinically relevant concentrations. Binding studies indicate a specific active site for volatile anesthetics on neuronal nicotinic receptors.2,19 Specific mutations on the M2 domains of the nicotinic receptor, which correspond to the α-helix segments that form the aqueous pore of the receptor’s ion channel, enhance the blocking action of anesthetics such as isoflurane and alcohols such as octanol.5 It is postulated that hydrophobic general anesthetic agents bind at a discrete site in the vicinity of the mutation loci. The more polar alcohols gain access to the same site preferentially after channel opening, suggesting that the binding site is within the channel itself. Inhibition of nicotinic receptors in skeletal muscle probably contributes to the ability of volatile anesthetics to enhance muscle relaxation. Actions at neuronal nicotinic receptors promote effects such as amnesia, hyperalgesia, and excitation observed at subanesthetic concentrations of volatile anesthetics and barbiturates.
The γ-aminobutyric acidA (GABAA) receptor has been implicated in the CNS depressant effect of most anesthetic drugs.4,30 Specific binding sites for benzodiazepines, barbiturates, other intravenous anesthetics, and volatile anesthetics have been described.4 Stimulation of these receptor sites increases the activity of GABA at its own separate site; many agents other than benzodiazepines can also open the GABAA Cl− channel in the absence of GABA. Hyperpolarization of the affected neuron inhibits neuronal activity. Glycine receptors constitute another group of inhibitory receptors that are activated by at least some general anesthetics (inhalation anesthetics, alcohols, thiopental, and propofol) in clinically relevant concentrations.
Excitatory receptors blocked by specific anesthetic agents include N-methyl-d-aspartate (NMDA), kainate, and α-amino-3-hydroxy-5-methyl-4-isoxazole propionate (AMPA) receptors. Ketamine and nitrous oxide13 selectively inhibit NMDA receptors, whereas barbiturates and certain inhalation anesthetics block AMPA and kainate receptors.4
In addition to the classic ligand-gated ion channels described previously, other ion channels may be involved in the actions of specific general anesthetics. Several types of 2-pore-domain K+ channels (identified by their acronyms TREK1, TREK2, TASK1, TASK3, and TRESK) are variably activated by inhalation anesthetics.4,6 These channels are responsive to intracellular second messengers and are believed to regulate background neuronal excitability and neurotransmitter release. Several types of Ca++ channels and Na+ are inhibited by clinical concentrations of drugs and may contribute an inhibitory influence on neurotransmitter release.