Chapter 38. Pharmacology of Inhalational Anesthetics

1. The most useful definition of "dose" for inhaled anesthetics is the partial pressure in alveolar gases, which is readily monitored in end-tidal gases.

2. All halogenated anesthetics break down, releasing carbon monoxide (CO) and heat when they contact desiccated alkaline chemicals, such as those in common CO2 adsorbents. Potential harm to patients from this breakdown can be avoided by proper use and maintenance of anesthesia equipment and by using less alkaline CO2 adsorbents.

3. The rate at which the alveolar anesthetic concentration (FA or Palv) approaches the inspired (circuit) concentration (FI or Pcirc) depends on minute alveolar ventilation (increased ventilation accelerates equilibration), cardiac output (increased output slows equilibration), and the blood-gas partition coefficient of the anesthetic (high solubility slows equilibration).

4. Nitrous oxide (N2O) diffuses into air-filled spaces in the body, causing expansion, increased pressure, or both.

5. The minimum alveolar concentration (MAC) is the alveolar concentration of inhaled anesthetic that blocks movement in half of subjects in response to a surgical incision. MAC is influenced by age, pharmacologic and physiologic factors (eg, temperature), and genetic factors.

6. MAC-awake is the alveolar concentration of anesthetic causing loss of response to verbal commands in half of subjects. Amnesia is produced by inhalational anesthetic concentrations lower than MAC-awake.

7. Awareness and explicit recall of intraoperative events is attributable to inadequate delivery of anesthetics for the patient's needs. Without preventive measures, awareness during anesthesia occurs in about 1 of 750 patients and may cause psychological disturbances leading to posttraumatic stress disorder.

8. All potent volatile anesthetics in current use decrease mean arterial pressure in a dose-dependent manner. Severe cardiovascular and respiratory depression can occur even at low volatile anesthetic concentrations in elderly, hypovolemic, or critically ill patients. Avoidance of these toxicities requires vigilant monitoring and anticipation of anesthetic requirements.

9. Volatile anesthetics may increase heart rate, both by a baroreceptor reflex in response to decreased arterial pressure and a direct vagolytic effect on the heart.

10. Volatile anesthetics tend to increase respiratory rate and decrease tidal volume and blunt ventilatory responses to hypercapnia and hypoxia.

11. Desflurane is very pungent, and its use can be associated with airway irritability, bronchoconstriction, and laryngospasm during induction. Among volatile anesthetics, sevoflurane causes the least amount of airway irritation.

12. Volatile anesthetics vasodilate cerebral vessels, increasing blood flow while reducing cerebral metabolic oxygen consumption. Cerebral vascular responses to altered pCO2 are maintained in the presence of volatile anesthetics. N2O increases cerebral metabolism.

13. Halothane undergoes the most hepatic metabolism of the inhaled agents. Whereas enflurane, isoflurane, and sevoflurane are also metabolized in the liver, desflurane and nitrous oxide are minimally metabolized. Oxidative metabolism of halothane and other volatile agents can induce a severe immune-mediated hepatitis.

14. All potent volatile agents may trigger malignant hyperthermia in susceptible individuals.

Experimentation with inhalation of gases and vapors for the purpose of obtunding the distress associated with surgery began in the 19th century. The administration of inhaled anesthetics spread rapidly after the successful public demonstration of ether anesthesia by William T.G. Morton on October 16, 1846, at Massachusetts General Hospital.

Inhaled agents were the sole means of reliably inducing general anesthesia until the development of intravenous (IV) delivery techniques and drugs. Inhalants continue to be used in a large fraction of general anesthetics because of their ease of use and predictable effects. The inhalational route of administration is almost always available, and the same route is used for drug removal. Thus, inhaled anesthetics require no metabolic clearance (indeed, their metabolism is associated with toxicity). Modern equipment for administration of inhaled anesthetics is simple and robust, providing an elegant method for inducing, maintaining, and reversing general anesthesia. Furthermore, monitoring inhaled anesthetic concentrations in end-tidal gases provides an estimate of drug concentrations in the circulating blood and central nervous system (CNS). This ability to assess drug concentrations in the body reduces pharmacokinetic uncertainty when determining how much inhaled vapor to administer.

This chapter describes the chemical and biophysical properties of commonly used inhaled anesthetics and relates these properties to the clinical pharmacology. Various properties of inhaled drugs influence their chemical stability, rate of pulmonary uptake, distribution to various body compartments, elimination, and metabolism (pharmacokinetics), as well as clinically important differences among their therapeutic actions and toxicities (pharmacodynamics). Understanding these pharmacologic relationships is crucial for the safe and effective delivery of inhaled anesthetics. An understanding of inhaled anesthetic pharmacology must begin with clear definitions of both the drug amount (dosage) and important effects (response).

The "dosage" of an inhaled anesthetic can be a confusing concept. The dosage of inhaled anesthetic documented in an anesthetic record is most frequently the delivered gaseous concentration (in percent of total gas) in the vaporizer outflow (ie, fresh gas). Dosage may also be defined as the inspired concentration of inhaled anesthetic in the breathing circuit or, analogous to injected drugs, as the amount absorbed by the body via the lungs. However, these definitions are of limited use because several factors, including ambient atmospheric pressure, fresh gas flow (FGF) rate, minute ventilation, and the rate of uptake into the blood via the lungs, determine how these various "doses" affect patients. The most useful and practical definition of dosage for inhaled anesthetic drugs is the partial pressure in alveolar gas, which is approximated in end-expiratory gas. The partial pressure of an inhaled drug is directly proportional to its fractional concentration in a gas mixture (Dalton's law) and is measured on an absolute scale of pressure (see Biophysical Properties of Inhaled Anesthetics, later).

The common effects that all general anesthetics reversibly induce are hypnosis (loss of perceptive awareness), amnesia (anterograde loss of memory), and ablation of movement in response to pain (inhibition of nociceptive reflexes).1 These therapeutic actions, which define the state of general anesthesia, are all mediated by the CNS (brain and spinal cord). Some anesthetics can provide additional therapeutic actions, such as analgesia, attenuation of autonomic reflexes, and protection of the heart and brain from ischemia and reperfusion. In addition, the nontherapeutic effects of anesthetics (side effects) must be considered because these often influence the choice of anesthetic drug and dosage, depending on the specific clinical setting.

The structures of various inhaled anesthetics, in 3 chemical categories, are depicted in Fig. 38-1. Three inhaled anesthetics (nitrous oxide [N2O], diethyl ether, and chloroform) were used in the 19th century. In the early 20th century, more inhaled anesthetics were identified, including ethylene, ethyl chloride, and cyclopropane. Of these, only N2O remains in use. The other early inhaled anesthetics in use up to 1950 were combustible, occasionally resulting in catastrophic outcomes for both patients and caregivers.2 The chemical and physical properties of the gaseous and volatile anesthetics currently available for clinical use (N2O, isoflurane, enflurane, halothane, desflurane, and sevoflurane) are summarized for comparison in Table 38-1.

Nitrous Oxide

Nitrous oxide is a simple linear inorganic compound that is in the gas phase at normal ambient temperature and pressure and is chemically stable. N2O has no odor or taste. The boiling point of N2O is –88.5°C. At room temperature (25°C), N2O condenses into a liquid at 745 psi (50 atm), making the storage and transport of large quantities in pressurized cylinders economical. A room temperature cylinder will display a pressure near 745 psi as long as liquid N2O remains. The pressure will decrease if the temperature of the liquid drops or if the tank is completely depleted of liquid and nearly empty. Pressure within the cylinder may decrease during rapid delivery of N2O as vaporization absorbs heat, cooling the tank and its contents. The amount of N2O remaining in the cylinder can only be determined by its weight, not by a pressure reading. N2O can support combustion, and the use of N2O during surgery with electrocautery or lasers may accelerate the fire if flammable materials ignite.

Nitrous oxide produces analgesia, yet it has a low potency as an anesthetic. It must be delivered at nearly 0.7 atm (530 mm Hg) to ablate awareness in half of patients, and preventing movement during an incision would require more than 1 atm in most patients. Therefore, N2O is frequently used in combination with other inhaled or IV anesthetic agents.

Halothane

In the mid-20th century, efforts to develop safer inhaled anesthetics focused on reducing flammability by halogenation (adding bromine; chlorine, and fluorine) of alkanes and ethers.3 Halothane is a halogenated alkane (CF3CHBrCl) and became clinically available in 1956. It has a pleasant, nonpungent odor and is tolerated well during inhalation. Halothane is a liquid at room temperature and 1 atm pressure, and its high vapor pressure (243 mm Hg at 20°C) is many times that needed to induce anesthesia. Clinically used concentrations of halothane are not flammable, but higher concentrations (in anesthesia machines) can ignite. Halothane is slightly unstable, decomposing in the presence of light and oxygen. To prevent photo-oxidative breakdown, halothane is stored in dark glass bottles containing 0.01% thymol. Newer halogenated volatile anesthetics do not require such chemical stabilizers. Halothane undergoes significant metabolism in the liver (Table 38-2).

Enflurane and Isoflurane

Enflurane and its isomer isoflurane were introduced to clinical practice in the 1970s. They are halogenated ethers that undergo much less metabolism and have faster onset and offset compared with halothane.

Enflurane is a halogenated ether (CHFClCF2OCHF2) and is more pungent than halothane but is reasonably well tolerated for inhalation induction. It is a clear liquid with a vapor pressure of 172 mm Hg at 20°C.

Isoflurane is a halogenated ether and an isomer of enflurane (CF3CHClOCHF2). It is a clear liquid with a vapor pressure of 238 mm Hg at 20°C. It is pungent and frequently stimulates coughing during inhalation induction.

Desflurane and Sevoflurane

Desflurane and sevoflurane, introduced in the 1990s, are ethers that are halogenated exclusively with fluorine. Both are less potent and less blood soluble than other halogenated anesthetics. Low blood solubility provides rapid pulmonary uptake and elimination, which is highly desirable in clinical settings during which rapid emergence from anesthesia is valued.

Desflurane (CF3CHFOCHF2) is extremely resistant to biodegradation. Desflurane has a high vapor pressure at 20°C (664 mm Hg) and boils at 22.8°C (73°F). To prevent boiling, desflurane is stored in special bottles with valves that only open when fitted into the filling port of a vaporizer. The unique desflurane vaporizer heats the anesthetic to 39°C to control its vapor pressure and delivery rate.4 Desflurane vapor is extremely pungent, and coughing and autonomic stimulation are frequently produced during inhalation induction.

Sevoflurane [(CF3)2CHOCH2F] is a clear liquid with a vapor pressure of 157 mm Hg at 20°C. Sevoflurane has a pleasant odor and low pungency, making it well tolerated for inhalation induction of anesthesia.

The concepts of partial pressure and partition coefficients are central to understanding how gases distribute among various compartments in the body.

Partial Pressure

Partial pressure is the pressure exerted by 1 component of a gas mixture, in which the sum of all the partial pressures equals the total pressure. For example, air contains about 79% nitrogen (fraction of N2 = FN2 = 0.79) and 21% oxygen (FO2 = 0.21), so at 1 standard atmosphere barometric pressure (PBar = 1 atm = 760 mm Hg), the partial pressure of N2, PN2 = 0.79 atm = 600 mm Hg and PO2 = 0.21 atm = 160 mm Hg. Near sea level, where atmospheric pressure is about 760 mm Hg, the fractional concentration of a gas differs insignificantly from its partial pressure (in atm), and the 2 terms can be used interchangeably. At high altitudes, where ambient pressure is lower, air contains the same fractional concentrations of nitrogen and oxygen, and their partial pressures are reduced relative to those at sea level. For example, in Denver (PBar = 630 mm Hg), PN2 = 500 mm Hg and PO2 = 130 mm Hg. Similarly, gaseous anesthetics such as N2O, when used at high altitudes, exert a lower partial pressure (and have a subsequently reduced anesthetic effect) relative to the same fractional concentration at sea level.

Vapor Pressure

Volatile anesthetic vapor pressure is the gaseous partial pressure at the liquid–gas interface, such as that in a vaporizer. Vapor pressure remains independent of total ambient pressure and is solely a property of the anesthetic agent and its temperature. Thus, in relation to carrier gases, the fractional concentration of a saturated volatile anesthetic in a vaporizer is higher at high altitude (Fig. 38-2). When working at ambient pressures that differ significantly from sea level, anesthetists must adjust the delivered oxygen concentrations and vaporizer settings appropriately.

Partial pressure is the driving force for the diffusion of gases across permeable barriers into other gases, liquids, or tissues. At equilibrium, the partial pressure of any gas is equal in all intercommunicating compartments within a closed system. In the operating room, intercommunicating compartments include the anesthesia machine, the breathing circuit, and the patient's body, which is further compartmentalized. The partial pressure of an inhaled anesthetic is directly proportional to its concentration in certain compartments that may be liquid phase (ie, blood) or tissue (eg, brain), where concentration is usually defined as the weight, volume, or moles of drug per volume of liquid or tissue. Importantly, various liquids (eg, blood or cerebrospinal fluid) and tissues may contain remarkably different anesthetic concentrations at the same partial pressure, depending on the solubility of the anesthetic gas in each liquid or tissue. The concentrations in 2 different compartments define a partition coefficient.

Partition Coefficients

A partition coefficient is defined as the ratio of concentrations of a drug in one compartment (gas, blood, or tissue) versus another intercommunicating compartment at equilibrium (ie, when the anesthetic partial pressure is equal in both compartments). Partition coefficients therefore have no units. Another useful concept is that the partition coefficient represents the number of volumes of the reference phase (eg, gas) that contains the same amount of anesthetic as the second phase (eg, blood), which we will call the effective volume (or relative volume) of the second phase (Fig. 38-3; Table 38-2). The effective volume concept is similar to the distribution volume of injected drugs when calculating metabolic or renal clearance and helps illustrate why drug equilibration in different tissues takes vastly different amounts of time. Both blood–gas partitioning and tissue–blood partitioning (see Table 38-2) determine the distribution of inhaled anesthetics within the body.

Blood–Gas Partitioning

The blood solubility of inhaled anesthetics is another term for the blood–gas partition coefficient (λb/g). At equilibrium, the concentration (eg, volume of drug per volume of gas, liquid, or tissue = v/v) of most inhaled anesthetics is higher in blood than in surrounding gases. That is, λb/g is greater than 1.0 (Table 38-2). Two notable exceptions are N2O and desflurane, which have the lowest blood solubility of the commonly used inhaled anesthetics. Blood–gas partitioning varies depending on temperature, hematocrit, and the lipid content of blood. The solubility of most gases, including inhaled anesthetics, increases as the temperature of a liquid such as blood decreases.5 Blood cells and plasma contain protein and lipids, which provide a higher capacity for anesthetics. Thus, hemodilution (decreased cell mass) reduces and hypertriglyceridemia increases the blood solubility of inhaled anesthetics. Elevated blood lipids after a fatty meal increase anesthetic solubility relative to that after fasting.6

Tissue–Blood Partitioning

In general, the solubility of inhaled anesthetics in tissues depends on the fraction of tissue that is lipid7 because most anesthetics are highly lipophilic. Thus, the more potent volatile anesthetics, which are also the most oil soluble (based on the Meyer-Overton correlation), tend to partition avidly into fatty tissues (Table 38-2).

A Multicompartmental Kinetic Model

The inflows and outflows between the anesthesia machine circuit and the patient's lungs, blood, and tissues can be depicted as a cyclical multicompartmental system wherein the anesthetic partial pressure in 1 compartment is the upstream partial pressure driving anesthetic into downstream compartment(s) (Fig. 38-4). These compartments have different phases: gas in the anesthetic circuitry and airspace of the lung, liquid in blood, and mixed liquid and solid in organs and tissues.8 The model depicted in Fig. 38-4 is complex and requires computer assistance to calculate how the anesthetic partial pressure in various compartments changes over time (for details, see the legend of Fig. 38-4). It is, however, relatively easy to understand how an anesthetic moves between 1 upstream compartment and 1 downstream compartment (a 2-compartment system).

Transfer of anesthetic gas from one compartment to another (eg, from a vaporizer to the breathing circuit) is proportional to both the bulk carrier flow (FGF in this example) and the partial pressure difference (eg, Pdelivered – Pcircuit). As more anesthetic gas is transferred, the partial pressure difference between the compartments becomes smaller, and the transfer rate slows. The time it takes a given compartment to equilibrate with upstream anesthetic partial pressure is determined by its volume (or effective volume) and the bulk flow carrying anesthetic to that compartment.

Equilibration between 2 compartments is quantitatively described by an exponential equation (Eq. 38-1), which is characterized by a time constant (τ). After each time constant, the difference between upstream and downstream concentrations is reduced by about 63%. A general equation describing this is

where t is the elapsed time measured from when the anesthetic is turned on during induction. The time constant (τ) is directly proportional to the downstream compartment volume (V): If the volume of the receiving compartment doubles, τ doubles and it takes twice as long for equilibration to occur. τ is inversely proportional to carrier flow (F): If the delivery flow doubles, τ halves and equilibration occurs twice as fast.

After 5 time constants have elapsed, anesthetic concentrations in upstream and downstream compartments differ by less than 1%. At this equilibrium state, the compartments have the same partial pressure of anesthetic, and no net anesthetic transfer between compartments occurs even though carrier flow continues.

Equilibration time is independent of the source partial pressure if this pressure is small (see the discussion of the concentration effect below). The time, however, to reach a target partial pressure in the downstream compartment (eg, to reach a specific depth of anesthesia) will be shorter if the upstream partial pressure is higher. A common practice is to deliver potent volatile anesthetics at 2 to 3 times the target partial pressure (overpressure) to reduce the induction time. When the desired depth of anesthesia has been achieved, the vaporizer is set to a lower concentration to prevent the toxic effects of overdosage.

The Inspired Partial Pressure of Anesthetic Gas

The example cited above can be used to illustrate how long it takes to "prime" a breathing circuit with anesthetic, such as is done before a single breath induction. The upstream source is the output from a vaporizer (Pdelivered), the drug carrier is FGF, and the breathing circuit volume is Vcirc. The delivery of anesthetic is determined by the product of the FGF and Pdelivered. The time needed to equilibrate the circuit with the delivered gas is determined by the ratio of Vcirc to FGF. High FGFs and low circuit volume, such as that in open-circuit delivery systems, enable rapid control of the anesthetic concentration in the circuit that is inhaled by the patient (Fig. 38-5).

The most common breathing circuit configuration is a circle system that allows rebreathing of exhaled anesthetic gases while removing carbon dioxide. These circuits have total gas volumes of 6 to 8 L. If Vcirc is 6 L and FGF is 6 L/min, then τ will be 1 minute, and it will take approximately 5 minutes for the anesthetic concentration in the circuit to closely approach the concentration delivered by the vaporizer. However, at FGF of 1 L/min, τ is 6 minutes, and it will take up to a half hour to fully equilibrate the circuit (Fig. 38-5A).

To illustrate the impact of additional compartments in our model, consider what happens if a patient is breathing circuit gases when the vaporizer is turned on. In this case, uptake of anesthetic into the patient's body reduces the rate of rise of the partial pressure of anesthetic in the circuit (Fig. 38-5B). These examples illustrate why a high FGF is most useful when rapid changes in inspired anesthetic concentration are needed, such as during induction and emergence. After the induction period, the rate of uptake of anesthetic into the patient slows, and if a rebreathing circuit is used, FGF can be reduced, usually in the range of 1 to 3 L/min. After FGF is reduced, modestly increasing the delivered anesthetic concentration will help maintain drug delivery to match the uptake of anesthetic into the body (Fig. 38-6). Reducing FGF has economic, environmental, and medical benefits. Lower FGF results in lower usage of inhaled gases, reducing costs as well as adverse environmental effects. The patient's airway humidity is also maintained, improving the clearance of bronchopulmonary secretions.

The extreme of low FGF is "closed circuit" anesthesia, wherein almost all inhaled gases are rebreathed and fresh gas is delivered with only sufficient oxygen to replace that metabolized by the patient (about 0.25 L/min for a typical normothermic 70-kg adult) and anesthetic vapor is added to match uptake and maintain the level of anesthesia. The practice of closed circuit anesthesia should be based on detailed knowledge of anesthetic gas uptake into patients (for a more thorough description, see Baum9).

The Alveolar Partial Pressure of Anesthetic Gas

The alveolar anesthetic partial pressure (Palv) is important because it is the upstream source or driving pressure that rapidly equilibrates with the blood, brain, and other highly perfused tissues (see below) and because its level can be monitored in end-expired gases. Thus, Palv represents the most useful definition of inhaled anesthetic "dose."

Transfer of Anesthetic from the Breathing Circuit to Alveoli

The transfer of inhaled anesthetic from the breathing circuit into the lungs depends on the minute alveolar ventilation rate and the difference between the partial pressures in the circuit (Pcirc) and the lung (Palv). Thus, hyperventilation enables the rapid adjustment of Palv. The impact of varying minute alveolar ventilation is illustrated in Fig. 38-7.

It is traditional to illustrate uptake with Palv normalized to Pcirc (inspired), which represents the case of high flow open-circuit anesthesia, wherein Pcirc ≈ Pdelivered. However, with a rebreathing circuit, as we have described, Pcirc/Pdelivered is not constant (Fig. 38-5). In this instance, a clearer illustration of how Palv varies over time is obtained by normalizing to Pdelivered, which is held constant in our uptake model calculations. The difference between open-circuit versus rebreathing models is illustrated in Fig. 38-7. Comparing the 2 panels, it is apparent that Palv increases more slowly with moderate FGF using a rebreathing circuit, but the impact of changing minute ventilation is similar in both cases. For simplicity, we have adopted the traditional open-circuit model to illustrate the impact of other physiologic variables on anesthetic uptake (for washout in these figures, we have normalized to either Pdelivered or Pcirc just before the agent is turned off).

Anesthetic Uptake into Pulmonary Blood

Pulmonary blood rapidly takes up alveolar anesthetic gases. The uptake into blood depends on blood solubility (the blood–gas partition coefficient, λb/g), the rate of pulmonary blood flow (usually similar to cardiac output), and the difference between the anesthetic partial pressure in the lung and that of mixed venous blood entering the lung. The more blood-soluble anesthetics produce a slower rise of Palv relative to the inspired concentration (Pcirc) because more vapor needs to be transferred into blood before the blood compartment is "filled" (Fig. 38-8). Stated another way, the effective blood volume is larger for the more soluble anesthetics (see Table 38-2).

Whereas increased cardiac output (pulmonary blood flow) slows the rise in Palv by more rapidly removing the anesthetic agent from alveolar gases, decreasing cardiac output will accelerate the rise in Palv. This effect is most significant for the highly blood-soluble anesthetics, for which the rate of rise of Palv is reduced more by vapor uptake into blood. Changes in cardiac output cause smaller changes in the rate of rise of Palv for the less blood-soluble anesthetics such as N2O and desflurane (Fig. 38-9).

The role of cardiac output in inhaled anesthetic uptake (and therefore inhalation induction) can seem counterintuitive. Increased cardiac output results in a more rapid delivery of anesthetic to the major sites of action in the nervous system, suggesting that anesthesia induction might be faster, not slower. The key to untangling this conundrum is to understand that during inhalation induction, Palv is the driving force for anesthetic drug entry into the blood and nervous system. Because Palv increases more slowly when cardiac output increases, partial pressure in the brain and spinal cord also increases more slowly. What happens to the "extra" anesthetic that is taken up via the lung at an elevated cardiac output? The brain and other highly perfused tissues (see below) equilibrate rapidly with Palv. However, muscle and fat tissue, which typically take hours to equilibrate with Palv, are absorbing most of the additional anesthetic agent.

In summary, the rate of equilibration between the alveolar partial pressure (Palv) and the inspired anesthetic partial pressure (Pcirc) depends on 3 factors: (1) Minute alveolar ventilation (as minute alveolar ventilation increases, the lung more rapidly equilibrates with the circuit), (2) cardiac output (as cardiac output increases, uptake from the pulmonary airspace into the blood and tissues slows the rate at which Palv rises), and (3) blood–gas partitioning (high blood solubility, or λb/g, increases uptake into blood, removing a higher fraction of anesthetic from the pulmonary airspace and slowing the rate at which Palv increases).

Distribution of Inhaled Anesthetics in Body Tissues

Mixed venous blood entering the pulmonary capillary bed rapidly equilibrates with the alveolar partial pressure of an inhaled anesthetic (Palv) so that blood exiting from the pulmonary vein has a partial pressure close to Palv. Anesthetic is then distributed to various tissues via systemic arterial blood. Delivery of anesthetic to each tissue is determined by its blood flow rate, its anatomic volume, and its tissue–blood partition coefficient (Table 38-2).

The effective volume (Veff) for uptake of anesthetic in a given tissue is the product of its anatomic volume and λtissue/blood. The time it takes for anesthetic partial pressure in a given tissue to equilibrate with Palv is decreased when tissue perfusion is high and increased when Veff is large (see Eq. 38-2).

Highly perfused tissues include the brain, spinal cord, kidney, liver, and heart, which comprise less than 10% of an adult body's mass while normally receiving about 70% of the cardiac output. Highly perfused tissues equilibrate within a few minutes with the arterial (alveolar) anesthetic partial pressure (Fig. 38-10, Table 38-2). As a result, the alveolar partial pressure of anesthetic, which can be measured in end-tidal gases, is usually close to the brain and spinal cord partial pressures (except when the anesthetic concentration in alveolar gas is changing rapidly).

Muscle represents a large anatomic volume (average, 35%-40% of body mass) and the muscle–blood partition coefficient for all inhaled agents ranges from 1.2 to 3.1. Because muscle normally receives only about 10% of cardiac output at rest, the anesthetic partial pressure in muscle rises slowly, and equilibration with Palv takes hours for most clinically used anesthetics (Table 38-2).

Fat receives about 13% of cardiac output and represents about 25% of body mass in the average adult, with considerable variation. In addition, the potent volatile agents partition highly into fat (λfat/blood; range, 27-51), so that the effective volume for uptake of volatiles into fat is extremely large (Table 38-2). For the volatile anesthetics, uptake into fat is so slow that equilibration of the partial pressure with Palv never occurs under normal clinical conditions. The only commonly used inhaled anesthetic with a low fat solubility is N2O (λfat/blood = 2.3), for which the equilibration time constant is about 2 hours.

The vessel-poor group of tissues include skin, ligaments, tendons, cartilage, and cortical bone. These represent about 14% of the average adult's body mass and receive less than 2% of cardiac output. As a result, their contribution to anesthetic uptake is negligible.

Other Factors Affecting the Uptake and Distribution of Inhaled Anesthetics

Intrapulmonary right-to-left shunts (perfused but nonventilated lung regions) do not participate in anesthetic uptake. They slow the rate of uptake into blood and make Palv increase more quickly, especially for the highly blood-soluble anesthetics. At the same time, pulmonary shunting results in an arterial admixture of mixed-venous blood with blood from gas-exchanging lung regions. As the fraction of pulmonary blood flow passing through right-to-left shunts increases, Part, the direct upstream anesthetic pressure source for the nervous system, decreases relative to Palv. For the highly soluble anesthetics, the increased Palv somewhat compensates for admixing, and Part increases only slightly slower than normal (Fig. 38-11). With less soluble agents, the impact of shunt on uptake from the lung (Palv) is smaller and there is less compensation for admixing.

Systemic arteriovenous (left-to-right) shunting causes the anesthetic partial pressure in mixed venous blood to increase rapidly, slowing the uptake of anesthetic agent from alveolar gas. Alveolar anesthetic partial pressure therefore increases more rapidly, compensating for the reduced uptake rate. Overall, arteriovenous shunting modestly increases the rate of increase of anesthetic partial pressures in alveoli and in the highly perfused tissues.

Systemic right-to-left shunting, such as that in the lung, causes downstream arterial admixing. Right-to-left intracardiac shunts, such as patent foramen ovale or patent ductus arteriosus (PDA) with elevated right heart pressures, result in regional Part lower than Palv. PDA right-to-left shunting may result in different anesthetic partial pressures delivered to the brain and spinal cord. The overall effect of shunting can be difficult to predict because total cardiac output may also increase to compensate for the shunt.

Anatomic and physiologic deadspace (ventilated but nonperfused lung regions) effectively reduces alveolar ventilation relative to minute ventilation. This results in slower anesthetic uptake into blood from perfused lung areas, particularly for highly soluble drugs. At the same time, the lack of anesthetic uptake from nonperfused deadspace results in a faster increase in Palv, which compensates for the slowed uptake.

Anesthetic-induced changes in cardiac and respiratory function affect the rate of anesthesia induction with volatile anesthetics. When spontaneous ventilation is maintained, the uptake of inhaled drug is reduced as anesthetic is absorbed, and minute ventilation decreases. Anesthetic depression of cardiac output is another dynamic factor altering both uptake and distribution of anesthetic, as discussed above.

Concentration Effect and Second Gas Effect

Nitrous oxide represents a special case in clinical anesthesia because it is often the major constituent of the inhaled gas mixture. As a result, uptake of N2O from the alveolar space into blood produces significant shifts of alveolar gas volume.10 As alveolar gas volume diminishes, the alveolar concentration of N2O is thus maintained (the concentration effect) as are the alveolar concentrations of other gases (the second gas effect). Consider a single breath of a gas mixture of 70% N2O, 29% O2, and 1% isoflurane near the start of an anesthetic when mixed venous blood does not contain N2O (Fig. 38-12). When half of the alveolar N2O is taken up into blood, the remaining alveolar gas volume will be 65% of the original volume with relative concentrations of 35 N2O:29 O2:1 Iso = 54% N2O:45% O2:1.5% Iso. Because of the reduction in alveolar gas volume, the alveolar N2O concentration (and partial pressure) is more than half its original concentration (if 100% N2O were inhaled, uptake would not reduce the alveolar concentration at all). The concentration effect therefore maintains the pressure gradient for uptake and increases the rate of rise of Palv for N2O toward the inspired concentration. The second gas effect is also illustrated in Fig. 38-12. With N2O uptake, the alveolar concentrations of both oxygen and isoflurane can become higher than the inspired concentration, accelerating their uptake into blood.11

Nitrous Uptake into Air Spaces in the Body

Inhalation of N2O at high concentrations leads to important effects on airspaces within the body. Airspaces usually contain mostly nitrogen (79% of air), which has a low solubility in blood (λb/g = 0.015) such that bulk transfer of trapped nitrogen gas cannot occur on the timescale of most anesthetics. N2O (λb/g = 0.45) is delivered to airspaces 30 times faster than nitrogen leaves. As N2O diffuses into the preexisting airspace, the volume or pressure in the airspace increases. In most cases, both airspace volume and pressure increase, but examples in which one or the other effect dominates are useful for illustration.

A compliant airspace, such as a small air embolus or pneumothorax, can increase its volume with minimal pressure change. It will continue to expand until the partial pressure of N2O within the airspace equals that in surrounding blood. Theoretically, at 50% inspired N2O, the preexisting volume of the gas bubble would double (half of its gas volume would then consist of N2O). Similarly, 67% N2O could lead to a tripling of volume, and 75% N2O could quadruple the volume (Fig. 38-13). Small venous air emboli may thus become easier to detect with Doppler and echocardiographic monitors but may create more physiologic problems by obstructing blood vessels.12,13 A small pneumothorax could also expand to compress mediastinal structures and impair oxygenation or create hemodynamic compromise. The rate of air-space expansion depends on the volume, geometry, and compliance of the airspace as well as the blood flow delivering N2O to it. Small venous air emboli may expand within seconds, and a pneumothorax can expand toward equilibrium in less than an hour; airspaces in the bowel may take considerably longer to expand.

N2O diffusion into noncompliant airspaces can also create clinical problems. In a noncompliant space, volume does not change, so pressure increases as N2O enters. For example, after intravitreal injection of sulfur hexafluoride (SF6) or perfluoropropane (C3F8) gases, N2O administration can result in rapid increases of intraocular pressure, compromising retinal blood flow.14,15 Intracranial air bubbles (after craniectomy or pneumoencephalography) and the inner ear also represent airspaces in noncompliant body compartments.

Other common inhaled anesthetics also diffuse into airspaces, but their low concentrations in clinical settings result in negligible impact on airspace volume and pressure. Xenon is an experimental anesthetic inhaled at high concentrations (60%-70%), and similar to N2O, is associated with airspace expansion.16

Elimination of Inhaled Anesthetics Via Ventilation

The removal of inhaled anesthetics via the lungs is essentially the reverse of uptake, and many of the factors affecting induction also affect elimination rates. High FGF promotes faster washout of anesthetic from the circuit, enhancing the gradient for removal from the lung (Fig. 38-5). High minute ventilation clears alveolar anesthetic, reducing Palv and providing a gradient for movement of anesthetic from the blood to the alveoli (Fig. 38-7). Highly blood soluble anesthetics are retained longer than insoluble drugs because the effective volume of blood is higher as λb/g increases (Fig. 38-8). Similarly, increasing the cardiac output retards recovery (Fig. 38-9).

Some aspects of anesthetic elimination do not mirror the uptake during anesthetic induction. The use of overpressure to accelerate anesthetic induction has no parallel during recovery because the delivered concentration or partial pressure of anesthetic agent cannot fall below zero. Also, before induction, whereas the inhaled anesthetic partial pressure in all tissues is zero, different tissues will typically have reached different partial pressures when anesthetic delivery is discontinued. In particular, the partial pressures in muscle and fat may be far lower than Palv after a brief anesthetic. Because an anesthetic will continue to be transferred into tissue as long as Palv is greater than Ptissue, these compartments may continue to absorb the drug even after delivery is discontinued. This effect is illustrated in Fig. 38-10, where fat continues to take up isoflurane from blood long after delivery is discontinued, until Palv is less than Pfat. This uptake helps reduce the anesthetic partial pressure within the highly perfused tissues and can shorten emergence time.

The longer anesthesia is maintained, the more drug is absorbed into muscle and fat, which are characterized by a slow uptake and release of agent. Thus, clearance of inhaled anesthetics is context sensitive: the longer the anesthetic, the slower the clearance. Theoretical modeling suggests that after 4 hours at 1 × minimum alveolar concentration (MAC) of various anesthetics, 99.9% brain elimination takes 33 hours for desflurane, 52 hours for sevoflurane, and 71 hours for isoflurane and that significant amounts of these drugs remain in the body for days after discontinuation of anesthetic delivery.17 If fat and muscle reach high partial pressures (after many hours for most potent agents), recovery from anesthesia may be significantly delayed. Obese patients and those with a higher than normal muscle mass are therefore at risk for slowed emergence after long exposures to soluble anesthetics.

Therapeutic Effects and Anesthetic Depth

Although general anesthesia results in a multitude of physiologic alterations, there is no consensus among clinicians and researchers as to which actions are essential to the state of general anesthesia.18 Our view is that the common desired outcomes of general anesthesia include hypnosis (loss of awareness), amnesia (loss of memory), and immobility (suppression of movement in response to pain). Some have argued that analgesia is also an essential component.19 Pain is unquestionably a critical consideration in perioperative care, but the potent sedative–hypnotic effects of most general anesthetics confound the assessment of pain. Hypnosis and amnesia are produced via drug effects on neural networks within the brain, and immobility is primarily mediated by the spinal cord. In short, the major therapeutic actions of the volatile anesthetics take place in the CNS. Beneficial and toxic effects of inhaled anesthetics on other physiologic systems (eg, cardiovascular function, respiration) can thus be regarded as secondary. In this section, we discuss various measurements of anesthetic effects in the nervous system, as well as their limitations.

Because general anesthesia is defined as the loss of normal responses to environmental stimuli, anesthetic depth is most rigorously defined by stimulus–response testing using stimuli that range from benign (eg, spoken commands) to noxious (eg, laryngoscopy or surgical incision).20 In addition, certain consistent pharmacologic effects of anesthetics that are stimulus independent are useful signs of anesthetic depth. Traditionally, both desired and undesired clinical effects have been associated with the various stages and planes of anesthesia introduced by Snow (1847)21 and modified by Guedel (1937).22 These descriptions (Table 38-3) were developed during the era of ether and chloroform anesthesia and are therefore not fully applicable to modern practice.

Minimum Alveolar Concentration, Minimum Alveolar Concentration Awake, and Minimum Alveolar Concentration Bar

Minimum Alveolar Concentration

In 1965, Eger introduced the concept of MAC as a stimulus–response measure of anesthetic potency.23,24 MAC is the alveolar concentration of inhaled anesthetic that prevents movement in half of subjects in response to a surgical incision. Thus, MAC is the equivalent of an ED50. (dose of a drug that produces an effect in 50% of subjects) inhibition of movement in response to a specific noxious stimulus. If different noxious stimuli (eg, varying point pressures or electric shocks) are used, the concentration of inhaled anesthetic required to suppress movement increases with stimulus intensity.25 Thus, MAC is most useful for comparing potency among different inhaled agents under the same conditions, with potency being inversely related to MAC. During measurement of MAC, equilibrium between the alveolar gas compartment and the CNS must first be established. Other drugs that modulate awareness (eg, benzodiazepines), pain sensation (eg, opioids), or movement (eg, muscle relaxants) cannot be present. MAC as originally defined depends on atmospheric pressure, but when agent concentration is expressed as a partial pressure, MAC becomes independent of ambient pressure. The MAC values of common inhaled agents in oxygen (Table 38-2) show that N2O is least potent followed by desflurane, sevoflurane, enflurane, isoflurane, and halothane. By definition, an exclusively inhalational anesthetic to a level of 1 × MAC will prevent movement in only 50% of patients. The ED95, which is roughly 1.3 × MAC, may be a more clinically useful value. The ED95 corresponds to 0.9% for halothane, 1.68% for isoflurane, and 1.88% for enflurane.

MAC is altered by age and physiologic, genetic, and pharmacologic factors. MAC decreases with age (Fig. 38-14). Standard MAC values are those for patients around age 40 years. MAC is highest within the first year of life (age 6-12 months) and decreases with advancing age.26,27

Physiologic factors such as temperature influence MAC. For each decrease in core temperature by 1°C, MAC decreases by 5%. Other physiologic extremes that affect CNS function (eg, hypoxia, hypercapnia, acidosis, hypotension) also decrease MAC.

Whether MAC is affected by gender is controversial. Elderly women require 26% less xenon than age-matched men. In young men and women, however, MAC for desflurane does not significantly differ.28 MAC is reduced in parturient women.29 Increases in either progesterone or endogenous opiates (endorphins) during pregnancy have been proposed to account for this decreased anesthetic requirement, but these theories have not been substantiated by experiment.

Genetic factors play a role in determining MAC. Mice of varying genomic backgrounds are differentially susceptible to volatile anesthetics such as halothane, isoflurane, and sevoflurane.30 In humans, patients with naturally red hair have a significantly higher desflurane MAC than other patients (Fig. 38-15).31 Ninety percent of the red heads tested had mutations reducing expression of the melanocortin-1 receptor gene. Genetically altered (knock-out) mice lacking the melanocortin-1 receptor gene also display a modest increase in MAC for desflurane, isoflurane, sevoflurane, and halothane.32

Pharmacologic factors alter MAC. The addition of 70% N2O reduces the anesthetic requirements of other inhaled agents by 55% to 70% (Table 38-2). MAC is an additive phenomenon when 2 or more inhaled agents are combined.33 Adjunct opiate or benzodiazepine administration reduces MAC. Whereas acute alcohol intoxication reduces MAC, chronic intake of alcohol or sedatives can increase MAC, a phenomenon known as cross-tolerance.34

MAC reflects the effects of inhaled anesthetics in the spinal cord. In animal models, MAC has been shown to be primarily dependent on anesthetic effects on the spinal cord and not the brain.35,36 Anesthetic-induced immobility is likely attributable to suppression of spinal motor neuron function, observed as a diminished Hoffmann's reflex (H-reflex) and F-wave amplitudes.37,38 Suppression of movement and H-reflex amplitude by sevoflurane follow similar dose-response relationships in humans.39

Minimum Alveolar Concentration—Awake

MAC-awake is the concentration of inhaled anesthetic that inhibits appropriate responses to spoken commands in half of patients.26 The ratio of MAC-awake to MAC is not consistent among inhaled agents. It is fairly constant for the halogenated volatiles (roughly 0.35 × MAC), and significantly higher for N2O,40 which likely reflects their different mechanisms of action (see Chapter 37).

Functional recovery from anesthesia is influenced by MAC-awake for the specific agent as well as its elimination kinetics. Therefore, the time required to awaken after an anesthetic, especially a long one, depends on what concentration of the anesthetic was used relative to its MAC-awake. If the inhaled concentration was 2 × MAC-awake, then only a 50% reduction in Palv will be needed before the patient awakens, which is usually rapidly achieved for most agents. If higher concentrations were used, then emergence may be more than proportionately slowed because the slower phases of agent elimination will have a more dominant effect (see Elimination of Inhaled Anesthetics via Ventilation).

Similar to MAC, MAC-awake is reduced in elderly adults, by hypothermia and by the presence of other drugs with hypnotic activity (ethanol, benzodiazepines, anticonvulsants, antidepressants).26 MAC-awake is also reduced by neuraxial blockade (spinal or epidural) despite intact cranial nerve function.41-44 This effect is believed to be caused by diminished ascending signals from the spinal cord, which stimulate cortical arousal via the brainstem. MAC-awake is decreased by high doses of opiates, but this reduction is smaller than that of MAC.45

MAC-awake (hypnosis) is associated with volatile anesthetics effects on the cerebral cortex. Gamma range (25-100 Hz) signals from different cortical regions become desynchronized during the transition from conscious to unconscious states.46,47 Synchrony in specific cortical circuits, particularly the thalamocortical network, is believed to be associated with higher cognitive function, and its interruption may account for loss of consciousness.48 Loss of "functional connectivity" among various cortical networks during anesthesia has been demonstrated by functional brain imaging studies.49,50

Low concentrations of inhaled anesthetics block memory more than awareness.40 When inhaled agents are given at MAC-awake, only a small fraction of "aware" experimental subjects (those responding to spoken commands) can recall events.51 Amnesia is likely caused by effects on the limbic network, including the amygdala and hippocampus. Lesions in the basolateral amygdala produce resistance to the amnestic effects of sevoflurane in animals.52 Volatile anesthetics both attenuate hippocampal activity and inhibit long-term potentiation of hippocampal synapses that is associated with memory formation.53,54

Minimum Alveolar Concentration for Blockade of Autonomic Responses

Minimum alveolar concentration for blockade of autonomic responses (MAC-BAR) is the alveolar anesthetic concentration that suppresses cardiovascular responses to surgical incision in half of patients.55 MAC-BAR is typically greater than MAC. For example, MAC-BAR for desflurane is 1.66 × MAC.55 Similar to MAC and MAC-awake, MAC-BAR is reduced by opiates.20

Limitations on Traditional Anesthetic Depth Measurements

Stimulus–response testing is usually impractical in clinical settings. Neuromuscular blocking drugs ablate motor responses to both painful (MAC) and benign (MAC-awake) stimuli. Techniques for maintaining a motor response capability are laborious. As a result, autonomic signs such as blood pressure, heart rate, diaphoresis, tearing, and pupillary responses are often the only accessible data for assessing depth of anesthesia. These signs are valuable but unreliable guides to anesthetic depth and are confounded by drugs and diseases that impair cardiac or autonomic nervous system functions.

Numerous factors affect anesthetic sensitivity (including unknown genetic factors), resulting in widely varying individual anesthetic requirements. Too little anesthetic puts patients at risk for intraoperative awareness (see below). Conversely, deep anesthesia for all patients is neither feasible nor advisable. Deep inhalational anesthesia in patients with cardiac disease, hypovolemia, and other critical illnesses predictably causes profound hypotension and organ hypoperfusion. Healthy patients may tolerate deep anesthesia, but it may result in a slow emergence and a high incidence of side effects. Recent research suggests that excessively deep general anesthesia may accelerate the pathogenesis of neurodegenerative diseases such as Alzheimer dementia56-58 and may be associated with increased late mortality.59 The delivery of sufficient but not excessive anesthesia while providing optimal surgical conditions represents a central challenge for anesthetists.

Electroencephalographic Measurement of Anesthetic Depth

Given that the modern practice of general anesthesia frequently includes neuromuscular blockade to provide immobility, narcotics to provide analgesia, and other drugs to control autonomic activity, the essential therapeutic effects of general anesthetics are hypnosis and amnesia (ablation of awareness and memory). These effects, which are mediated in the brain, are also the most difficult to assess clinically. Monitors that analyze electrical signals from the brain can provide anesthetists with more data to individually titrate the depth of general anesthesia to achieve these endpoints. In addition, titration to individual needs has been shown to reduce anesthetic dosage, resulting in faster emergence in some settings.60,61

A number of techniques based on electroencephalography (EEG) have been developed and used. Fourier transformation of raw EEG data enables the derivation of median power and spectral edge frequencies.62 Other EEG analyses assess bispectral phase relationships, burst suppression, and entropy. The bispectral index (BIS; Covidien, Boulder CO) is a proprietary algorithm based on burst suppression, near-burst suppression, beta-band power, and phase relationships between delta- and theta-waves.63 The patient state index (PSI; Physiometrix, Inc., N. Billerica, MA) is based on EEG component relationships between the frontal and occipital regions.64 Entropy monitors (eg, S/5, General Electric [Datex-Ohmeda], Helsinki, Finland) analyze entropy (randomness in frequency and phase relationships) using both EEG and frontal electromyography (EMG).65 Some of these monitors are now available for clinical use to provide additional information and help prevent undesired side effects such as awareness during general anesthesia. For most patients, EEG indices such as BIS and spectral entropy correlate with the alveolar concentration of volatile anesthetics (Fig. 38-16).66 Monitors based on stimulus–response are also being developed for clinical use. Auditory evoked potentials can be used to assess anesthetic-induced unconsciousness and may be used in conjunction with other EEG parameters.67 A recent innovation is the use of transcranial magnetic stimulation in conjunction with EEG to detect loss of consciousness.68 Nonetheless, all of these monitors have considerable limitations. Further research in identifying neural correlates of consciousness is needed to improve the utility of monitors that assess the depth of anesthesia.

The interindividual variability in sensitivity to general anesthetics and the difficulty in clinical assessment of anesthetic depth inevitably result in excessive or inadequate depth of anesthesia in some patients. Inadequate anesthesia can result in awareness and explicit recall of intraoperative events, a problem that is a subject of clinical, public, and scientific attention.69

Incidence of Awareness with Recall during General Anesthesia

Large studies in both Sweden70 and the United States71 have used multiple postoperative interviews (eg, Table 38-4) to investigate intraoperative awareness. In a total of 31 360 combined patients, 40 cases of "definite" and "probable" awareness with recall were identified, an incidence of 0.13%. Given that 30 to 50 million general anesthetics are administered in the United States, the number of patients experiencing awareness with recall is estimated at around 50 000 per year.

Patient Reports and Posttraumatic Stress Disorder after Intraoperative Awareness

Spontaneous reports by patients of intraoperative awareness are rare, highlighting the need to using a structured interview that probes for these events.72 Patients report a variety of intraoperative experiences after general anesthesia. A high proportion of these experiences are vague and dreamlike, but others are explicit recollections of intraoperative events. Explicit recollections may include hearing conversations among operating room personnel; pain associated with intubation or surgery; and extreme anxiety, particularly in paralyzed patients. These experiences can lead to postoperative psychological problems. In the most severe cases, patients may develop posttraumatic stress disorder (PTSD). PTSD is characterized by recurrent episodes of anxiety, irritability, anger, and vigilance, often associated with flashbacks or nightmares, avoidance of cues related to the trauma, and sleep disturbances.73 The incidence of PTSD after intraoperative awareness is not known, but it likely depends on the length of the awareness episode and the presence of pain, anxiety, and preexisting psychological problems. In a small study of 16 post-awareness patients, more than 50% developed symptoms of PTSD requiring psychotherapy.74

Risk Factors for Awareness during General Anesthesia

Factors that contribute to intraoperative awareness include the use of muscle relaxants, high individual requirements for anesthetics, surgical situations during which "light" anesthesia is typically used, and errors by the anesthesia care team (Fig. 38-17).75 Awareness with recall has been reported in patients who were anesthetized without muscle relaxants, but the incidence is much higher when muscle relaxants were given.76 Furthermore, severe psychological problems seem to develop primarily in patients who experience "awake paralysis" with anxiety and pain.77 Patients receiving chronic therapy with sedatives, opiates, and some anticonvulsants can develop tolerance (resistance) to anesthetic drugs and require higher than normal anesthetic doses. Pharmacogenomic factors are likely to play a role, and patients with either a personal or a family history of awareness during general anesthesia may be at increased risk. Surgical procedures associated with high incidences of intraoperative awareness include major trauma, thoracoabdominal, and cardiac surgery. In these settings, hemodynamic instability often precludes the use of sufficient depth of anesthesia to prevent awareness, resulting in increased incidences of awareness ranging from 1% up to 40%.78,79 Caesarean section under general anesthesia, during which a "light" anesthetic is often used to reduce depression of the newborn, is also associated with a high incidence of awareness (0.4%).80 Inadequate dosing, misuse of equipment, and other human errors can also lead to patient awareness during general anesthesia. Many cases of awareness appear to occur during prolonged attempts at airway intubation when the hypnotic effect of IV induction agents wears off while patients remain paralyzed with a muscle relaxant. Other reports have described awareness when anesthetic vaporizers were empty or when drug pumps were not delivering proper doses of hypnotic agents. Reviews of awareness during general anesthesia incidents report that most cases were preventable.81

Monitoring and Prevention of Intraoperative Awareness

There is evidence that 2 monitoring strategies may be effective at reducing the incidence of intraoperative awareness when inhaled anesthetics are used to produce hypnosis. These are based on 2 measurements, the EEG-derived BIS and end-tidal anesthetic partial pressures, both known to correlate with depth of hypnosis (perceptive awareness).

The BIS is the only EEG-based monitoring technique that has been extensively evaluated for its ability to reduce the incidence of awareness. Myles et al82 conducted a prospective, randomized, double-blinded, multicenter trial of 2463 patients at high risk of awareness with recall who were assigned to either "standard care" or BIS-guided anesthesia groups. Postoperative interviews identified 2 patients (0.16%) who experienced awareness with recall in the BIS-guided group and 11 (0.9%) in the high-risk control group. Of note, use of the BIS monitor in this study did not reduce the total incidence of intraoperative experiences reported by patients (including reports judged as "possible awareness" or "no awareness") but only those with "confirmed" awareness with recall. A number of confounding factors can also affect the BIS, reducing its ability to accurately gauge depth of anesthesia. These factors include electrical artifact from electrocautery devices, physiologic alterations such as hypoglycemia, low-voltage EEG caused by genetic variation or drugs, and neurologic abnormalities such as Alzheimer disease.83 A common confounding factor that can alter BIS and spectral entropy values is EMG (muscle) activity, which can be reduced with neuromuscular blockade.84,85 Furthermore, appropriate interpretation and intervention are required for effective use of these monitors. In Sebel's et al's incidence study,71 BIS monitors were used in about 40% of the patients, yet this group of patients did not have a reduced incidence of awareness reports. Finally, the value of BIS monitoring has been validated only for propofol and volatile anesthetics. When ketamine and N2O are present, the relationship between BIS value and perceptive awareness can be significantly altered.86

End-tidal anesthetic gas (ETAG) monitoring, which is available in most modern anesthesia systems, appears to be also useful for reducing the incidence of intraoperative awareness. Avidan et al87 conducted a prospective randomized trial that assessed explicit awareness in high-risk patients. The incidence of explicit awareness was similarly low (0.2%) in patients whose anesthetics were guided using BIS monitors and a comparison patient group who received ETAG concentration-guided anesthesia (0.7 – 1.3 × Age-adjusted MAC). However, there was no "standard care" control group included in this study, and thus, additional research is needed to verify that ETAG monitoring is effective at reducing the incidence of intraoperative awareness. A clear limitation of ETAG is that it is not applicable or reliable for patients receiving total IV anesthesia or a combination of IV infusions and inhaled anesthetic agents.

Management of Intraoperative Awareness

Prevention of intraoperative awareness with recall begins with recognizing and addressing known risk factors. Caregivers must be familiar with their equipment and its proper use. Amnestics, such as benzodiazepines, can reduce the incidence of awareness with recall and should be considered as premedication for most patients and when "light" anesthesia is unavoidable. Muscle relaxants should be used judiciously to provide adequate relaxation for surgery or avoided whenever possible. N2O-relaxant anesthesia should be supplemented with volatile anesthetics or IV hypnotic agents to maintain adequate hypnosis (˜2 × MAC-awake). When intubation is delayed or multiple attempts are required, consideration should be given to supplementing the induction bolus with additional inhaled or IV hypnotic agents. All patients should be routinely informed of the risk of awareness during general anesthesia. High-risk patients should be monitored using ETAG analysis or an EEG-based monitor when feasible and appropriate. If adjunct monitoring is used, it should be used according to manufacturer's guidelines, and indications of potentially inadequate anesthesia on the monitor should be treated with an increased dose of hypnotic agents unless other circumstances make this inadvisable.

Ideally, in the postoperative period, all patients should be interviewed and asked specific questions (Table 38-4) to assess for intraoperative awareness. When awareness is detected, the patient should be treated for this occurrence like any other complication that results in harm or potential harm to patients. A thorough debriefing of the patient is warranted, and the details of the patient's experiences should be documented in the medical record. All personnel involved in the surgery should be informed, and patient recollections should be corroborated as thoroughly as possible. The incident should also be reported to the departmental Quality Assurance Committee and possibly to the institutional legal counsel. Patients reporting awareness during anesthesia should be reassured that their experience is valid and provided with an explanation of why and how the intraoperative awareness occurred. Expressing sympathy for the patient's suffering and apology if indicated can help maintain a therapeutic relationship, and the patient should be offered professional psychological evaluation and therapy for his or her symptoms. Maintaining contact with the patient with daily visits or phone calls may facilitate their recovery and can reduce malpractice risk.

In this section, we summarize both the therapeutic and toxic secondary effects of inhaled anesthetics on important physiologic systems.

Central Nervous System

Cerebrovascular and Metabolic Effects

The effects of anesthetics on cerebral metabolism and vascular tone are summarized in Table 38-5. Normal cerebral blood flow (CBF) is tightly autoregulated and coupled with cerebral metabolic demands. Volatile anesthetics interrupt this autoregulation by acting as direct vasodilators of the cerebral vasculature.88 Volatile anesthetics dependently increase middle cerebral artery blood flow velocity, with sevoflurane causing less vasodilation than isoflurane or desflurane.89,90 Both K+ channels91 and neuronal nitric oxide synthase92 have been linked to these direct actions of anesthetics.

Volatile anesthetics also decrease the cerebral metabolic rate (CMR) of oxygen consumption.18,88 Cerebral metabolism is decreased by halothane, isoflurane, sevoflurane, and desflurane.93 The extent to which blood flow and metabolism are altered depends on the choice of agent. For example, halothane may increase CBF by almost 200% while reducing the CMR by only 10%.94 Isoflurane, by contrast, increases CBF by about 20% while reducing CMR by 45%.95,96 Volatile agents also partially uncouple flow–metabolism relationships and carbon dioxide reactivity.97,98 These agents have not been shown to increase intracranial pressure (ICP) in normocapnic adults undergoing supratentorial brain tumor resection without a preoperative midline shift.99 In children, isoflurane, sevoflurane, and desflurane have all been shown to increase ICP.100 N2O is distinguished from the potent volatile anesthetics in that it increases CBF and CMR and may increase ICP.101,102 This is partly attributable to its sympathomimetic effects.103

Electroencephalographic Effects

Volatile anesthetics alter the cortical EEG, decreasing high-frequency (gamma band) activity and increasing slower frequencies. Quantitative EEG analysis demonstrates that power shifts anteriorly.46 In comparison, N2O has little effect or increases the frequency of the cortical EEG.104-106 Indeed, EEG-based monitors do not reliably detect the hypnotic effects of N2O alone or in combination with other anesthetics.107,108 Deep volatile anesthesia (1.5-2 × MAC) leads to burst-suppression or isoelectric EEG patterns. Although all volatile anesthetics have been used to suppress seizures, both enflurane and sevoflurane have been reported to augment epileptic brain activity and may induce EEG patterns associated with epilepsy in normal patients.109-111 Enflurane and sevoflurane may be useful during cortical mapping for ablation of seizure foci but should otherwise be avoided in patients with seizure disorders.

The actions of anesthetics that are thought to produce the state of general anesthesia are discussed briefly above in this chapter and in more detail in Chapter 44. The long-term effects of anesthetics on cognitive function are reviewed in Chapter 84.

Cardiovascular System

The effects of volatile anesthetics on the cardiovascular system are of paramount clinical importance intraoperatively, particularly in patients at risk for end-organ ischemia. The effects are summarized in Table 38-6.

Mean Arterial Pressure

Volatile anesthetics decrease mean arterial pressure (MAP) in a dose-dependent manner93 by direct vascular and autonomic nervous effects. Volatile anesthetics reduce MAP by decreasing systemic vascular resistance, increasing vascular compliance (the change in circulatory volume with changes in pressure), and inhibiting myocardial contractility. Halothane depresses myocardial contractility more than other volatile anesthetics. Whereas isoflurane and desflurane decrease systemic vascular resistance, halothane, isoflurane, and sevoflurane increase arterial compliance112 (Table 38-7). Desflurane has been found to stimulate sympathetic nervous system activity (especially with rapid increases in vapor concentration), which may account for its minor effect on vascular compliance.113 N2O increases sympathetic activity and systolic blood pressure114 and thus counters the vasodilatory and hypotensive effects of coadministered volatile agents.115

Heart Rate

Volatile anesthetics increase heart rate, both by a baroreceptor reflex in response to decreased arterial pressure and by a direct vagolytic effect on the heart. In dogs, desflurane increases heart rate the most followed by sevoflurane, isoflurane, and halothane.116 Rapid increases in the alveolar concentration of desflurane are associated with both an increased heart rate and blood pressure,117,118 which is attributed to stimulation of sympathetic activity.113 The minimal effect of halothane on heart rate is associated with its inhibition of the baroreceptor reflex.119 N2O is associated with only modest transient increases in heart rate,120 which reflect its sympathetic effects and preservation of MAP.

Myocardial Contractility

Volatile anesthetics cause myocardial depression, partly by inhibiting calcium ion influx in the myocardium.121 Halothane causes a marked dose-dependent inhibition of myocardial contractility and reduces cardiac output.122,123 Isoflurane, sevoflurane, and desflurane have lesser effects on myocardial contractility. In dogs, all of these anesthetics depress myocardial contractility, delay cardiac chamber relaxation, reduce chamber stiffness, and impair left atrial–left ventricular coupling.124,125 N2O administration also decreases intracellular calcium levels and depresses myocardial contractility.126,127 Thus, although N2O minimally affects blood pressure and heart rate, it has negative inotropic effects similar to those of potent volatile agents.

Cardiac Rhythm and Conduction

Volatile anesthetics affect the function of cardiac ion channels, increasing the risk of arrhythmias.128 Whereas halothane sensitizes the heart to the arrhythmogenic effects of epinephrine,129 isoflurane, desflurane, and sevoflurane do not.130,131 Halothane, isoflurane, desflurane, and sevoflurane all prolong the QT interval.132,133 In dental outpatients, halothane produced more ventricular arrhythmias than isoflurane.134,135 In pediatric dental outpatients receiving halothane anesthesia, 48% had ventricular dysrhythmias compared with 8% to 16% of those who received sevoflurane.136 Furthermore, sevoflurane-induced arrhythmias were primarily single supraventricular ectopic beats, but ventricular tachycardia was observed in 12% of patients receiving halothane. N2O may induce atrioventricular junctional rhythms137 and can lower the threshold for epinephrine-induced arrhythmias in conjunction with halothane.138

Coronary Artery Perfusion

Isoflurane is a coronary vasodilator and can thus potentially induce "coronary steal," a diversion of blood flow away from fixed stenotic lesions.139 The clinical significance of this phenomenon appears to be mostly theoretical. Isoflurane has been observed to be safe in patients with coronary artery disease as long as adequate perfusion pressure is maintained. Isoflurane also provides beneficial effects via ischemic preconditioning.140 Desflurane and sevoflurane do not cause coronary steal.141,142

Respiratory System

Ventilation

Volatile anesthetics depress respiration through both central medullary and peripheral muscular effects. In general, inhaled anesthetics decrease tidal volume and increase respiratory rate. Halothane, isoflurane, desflurane, and sevoflurane dose dependently reduce tidal volume. The concomitant increase in respiratory rate is more pronounced with halothane, desflurane, and sevoflurane than with isoflurane.141,143-145 Desflurane maintains minute ventilation with compensatory tachypnea up to alveolar concentrations of 1.6 × MAC. Nonetheless, alveolar ventilation is reduced by all volatile anesthetics, resulting in an increased PaCO2. N2O also causes tachypnea and decreased tidal volume but alone causes minimal changes in PaCO2. Ventilatory depression is additive when N2O is administered in combination with other inhalational agents.146

Factors that contribute to hypoxia and hypercarbia during inhalational anesthesia include hypoventilation, atelectasis, airway closure, decreased functional residual capacity, and ventilation/perfusion (V/Q) mismatch.147 Volatile anesthetics blunt hypoxic and hypercarbic respiratory drive, increasing the risk of severe hypoxia and hypercarbia in spontaneously breathing patients.148-150 Depression of hypoxic and hypercarbic ventilatory drives occurs even at subanesthetic concentrations. N2O blunts the respiratory drive to hypoxia and hypoventilation, but its clinical effects are minimal because of its low potency.151 Rapid elimination of N2O from blood after discontinuation dilutes alveolar gases, and if supplemental oxygen is not supplied, it can lead to diffusion hypoxia.152,153

Hypoxic Pulmonary Vasoconstriction

Hypoxic pulmonary vasoconstriction (HPV) is a pulmonary vascular mechanism that diverts blood flow away from poorly ventilated areas of the lung, minimizing V/Q mismatch. The mechanisms underlying HPV are not yet fully understood, but cyclooxygenase, calcium channels, and potassium channels appear to be involved.154 N2O has been shown in vivo to inhibit HPV.155 In isolated lung models, halothane, isoflurane, sevoflurane, desflurane, and N2O all inhibit HPV dose dependently.156-158 In many animal studies, however, clinical concentrations of halothane, isoflurane, sevoflurane, and desflurane have not been shown to inhibit HPV.159-162 In a few studies of chronically instrumented dogs, isoflurane was found to attenuate HPV.163 The conflicting data regarding isoflurane effects on HPV may be attributable to different flow conditions in the pulmonary vasculature during different experiments.

Bronchial Tone

Most volatile anesthetics, including halothane, enflurane, isoflurane and sevoflurane, are bronchodilators that decrease respiratory resistance and increase dynamic compliance (Table 38-8).164 At equi-MAC concentrations, respiratory resistance is reduced by sevoflurane > halothane > isoflurane.165 In contrast to the other volatile anesthetics, desflurane is associated with either no bronchodilatory effects or bronchoconstrictor effects at higher concentrations.166 Inhalational induction with desflurane is associated with coughing and an increased risk of laryngospasm, which is likely caused by its pungency. Sevoflurane causes the least amount of subjective airway irritation and is well tolerated during inhalational induction.167 N2O has no effect on respiratory resistance.165

Hepatic and Renal Systems

Liver

Volatile anesthetics reduce overall hepatic perfusion by altering portal venous or hepatic arterial inflows. Most agents decrease portal venous flow. Isoflurane has the least overall effect on hepatic perfusion because of compensatory increases in hepatic artery flow.168 In contrast, halothane causes hepatic artery vasoconstriction and decreases overall hepatic blood flow,169 hepatic oxygen delivery, and hepatic vein blood oxygen saturation.170 N2O causes minimal circulatory effects on the liver.

Transient modest increases in liver enzymes are common after exposure to volatile anesthetics (halothane, desflurane, sevoflurane, and isoflurane). This effect is independent of surgical intervention and is almost always clinically insignificant.171,172 The increase in liver enzymes after halothane may be caused by anaerobic reductive metabolism that generates free radicals.173 More severe hepatic injury (halothane hepatitis) is described under Breakdown of Inhaled Anesthetics and Toxicity of By-products.

Kidney

Volatile anesthetic agents cause dose-dependent decreases in renal blood flow, glomerular filtration rate, and urine output, which can be minimized with preoperative hydration.174-176 Autoregulation of renal blood flow is preserved during halothane anesthesia.175 Decreases in renal blood flow during inhalational anesthesia reflect a reduction in effective circulating volume secondary to increased vascular capacitance. Anesthetics do not directly stimulate antidiuretic hormone (ADH) release, but diminished urine production is associated with an increase in ADH because of surgical stress.177 N2O causes minimal effects on renal blood flow and function. Nephrotoxicity can be associated with breakdown of volatile anesthetics (see Breakdown of Inhaled Anesthetics and Toxicity of By-products).

Other Systemic and Adverse Effects

Muscular System

Volatile anesthetics potentiate neuromuscular blockade via direct effects on the neuromuscular junction.178 The muscle relaxant effect of volatile agents is stronger than that of IV anesthetics; N2O has no relaxant action. In isolated diaphragmatic muscle, enflurane and sevoflurane enhance fatigability at high concentrations.179,180 This may be mediated in part by cyclic adenosine monophosphate because administration of the phosphodiesterase inhibitor olprinone attenuates the effects of volatile anesthetics.

Malignant Hypothermia

Malignant hypothermia (MH) is a life-threatening clinical myopathy triggered by all potent volatile anesthetics and succinylcholine. It is a genetic disorder associated with autosomal dominant transmission of genes encoding mutant forms of the skeletal muscle calcium release channel (the ryanodine receptor protein or RyR1).181 Upon exposure to the triggering drugs, patients with MH develop an exaggerated increase in intracellular calcium, resulting in sustained skeletal muscular contracture, which is not inhibited by neuromuscular blocking agents. Sustained contracture produces a hypermetabolic state, leading to increased CO2 production and eventually hyperthermia. This condition is fatal unless treated aggressively and rapidly with IV dantrolene.

Methionine Synthase Inhibition by Nitrous Oxide

Nitrous oxide irreversibly oxidizes the cobalt atom of vitamin B12, which inhibits the cobalamin-dependent enzyme methionine synthase. This pathway is essential for homocysteine breakdown, nerve myelination, methyl substitutions of neurotransmitters, and DNA synthesis. Most patients are unaffected by temporary inactivation of vitamin B12 after N2O administration. Nonetheless, this reaction can be clinically significant in patients with poor nutrition, preexisting vitamin B12 deficiency, or other metabolic diseases that converge on the same metabolic pathways. N2O exposure may lead to "anesthesia paresthetica," which is characterized by paresthesias, ataxias, and poor manual dexterity.182 Widespread neuronal damage, status epilepticus, and death were reported in 1 infant with a preexisting deficiency of 5,10-methylenetetrahydrofolate reductase, an enzyme in the methionine synthetic pathway, after exposure to N2O.183

Endocrine Function

Halothane and enflurane impair glucose tolerance in animal models by reducing both insulin secretion and receptor sensitivity. Isoflurane has been shown to increase endogenous glucose production and decrease glucose utilization.93,184 The selection of anesthetic may also play a role in modulating the stress responses to surgery. In a study of 20 women requiring laparoscopic surgery for ovarian cystectomy, sevoflurane was associated with less increase in cortisol and adrenocorticotropic hormone levels than isoflurane.185

Immune Function

Potent inhalational anesthetics can alter immune cell functions in various ways, which in turn may affect recovery from surgery and tumor viability. Halothane depresses the neutrophil oxidative response to inflammatory mediators of infection; this effect is smaller with desflurane, sevoflurane, and isoflurane.186 Sevoflurane impairs transcription factors in human lymphocytes, which may reduce the inflammatory response.187 Volatile anesthetics can also induce apoptosis in human T cells in vitro188 and may affect cytokine function.189 However, clinical studies suggest that the impact of anesthesia on immune function is transient and of little clinical significance.190

Ischemic Preconditioning

Volatile anesthetics induce cellular responses that protect against ischemia and biochemical stress mediators. For example, sevoflurane decreases markers of myocardial and renal damage after coronary artery bypass grafting.191 Animal models of renal ischemic injury demonstrate differential preconditioning, with desflurane demonstrating less protective effects against tubular necrosis than sevoflurane, isoflurane, or halothane.192 Volatile anesthetics can protect against neural ischemia in animal models through pathways involving both nitric oxide metabolism193 and potassium channels.194

Genotoxicity and Teratogenicity

Halogenated hydrocarbons and ethers can cause DNA damage. Halothane and isoflurane produce genotoxicity in proliferating blood lymphocytes in vitro; sevoflurane does not appear to be cytotoxic in an animal model.195 The potential for anesthetic-induced teratogenicity has been studied. Prolonged N2O exposure is teratogenic in a number of embryonic animal models.196 N2O inhibition of vitamin B12–dependent DNA synthesis is the likely cause. Because of potential genetic damage after chronic exposure to inhaled anesthetics, it is currently recommended that operating room air contain less than 25 parts per million (ppm) of N2O and less than 2 ppm of halogenated anesthetics. Under these conditions, there is little evidence of significant risk from workplace exposure to inhaled anesthetics.197 Furthermore, approximately 75 000 pregnant women undergo nonobstetric surgery each year. Although there was once controversy over the effects of surgery and anesthesia in this population, volatile anesthesia appears to be safe for both the woman and the fetus.198

Inhaled anesthetics represent a class of drugs that are eliminated largely via nonmetabolic pathways—for the most part, they leave the body as they entered, unaltered and via ventilatory gas exchange. Undesirable effects directly associated with anesthetics are summarized above, but others are caused indirectly by chemical decomposition of inhaled anesthetics into toxic by-products. The breakdown of volatile anesthetics into potentially harmful chemicals can occur in the presence of CO2 adsorbents or via enzymatic biotransformation in the body. In general, greater breakdown of inhaled anesthetics leads to greater toxicity.

Nonmetabolic Decomposition of Inhaled Anesthetics

Although inhaled anesthetics are chemically stable under normal storage conditions (including within vaporizers), decomposition can occur under certain environmental conditions.

Sevoflurane Breakdown and Compound a

When sevoflurane contacts CO2 adsorbents containing a strong base (Baralyme and Sodalime), chemical decomposition occurs, releasing volatile breakdown products.199 The major degradation product, compound A (fluoromethyl-2-2-difluoro-1-[trifluoromethyl] vinyl ether), was shown to cause renal injury and death in rats when inhaled at high levels.200 Renal injury is both dose and time dependent with a threshold for detectable injury in laboratory animals of 150 to 300 ppm hours. Nonetheless, in human volunteer and clinical studies, blood urea nitrogen (BUN) and creatinine levels remain unchanged after exposures that sometimes exceed 300 ppm hours. It is thought that humans sustain less renal injury than rats after compound A exposure because of lower levels of renal cysteine conjugated β-lyase enzyme activity.201,202 Special laboratory markers for subtle renal tubular damage in humans are elevated after 300 ppm hours of compound A exposure, and these normalize within a few days.203 In clinical settings, the inhaled concentration of compound A is proportional to the sevoflurane concentration and inversely related to fresh-gas flows. Whereas low gas flows allow compound A to accumulate in the breathing circuit, high gas flows wash out compound A with waste gases. At FGFs of 2 L/min or higher, concentrations of compound A are low enough that the conservative exposure threshold of 150 ppm hours is unlikely to be reached. Sevoflurane package labeling guidelines should be heeded. Clinical studies using standard renal function tests demonstrate that sevoflurane is no more harmful than isoflurane when administered with low FGFs to patients with preexisting renal disease.204

Volatile Anesthetic Breakdown and Desiccated Carbon Dioxide Adsorbents

All of the halogenated volatile anesthetics degrade in the presence of dry alkaline CO2 adsorbents in rebreathing circuits. (Only sevoflurane breaks down in the presence of moist adsorbent.) Decomposition in the presence of dry CO2 adsorbents releases carbon monoxide (CO), formaldehyde, methanol, and heat. These exothermic reactions have resulted in the ignition of breathing circuit components205 and acute respiratory distress syndrome in patients206 and can lead to significant carboxyhemoglobin levels in patients.207 These problems are avoidable and only arise when the CO2 adsorbent is desiccated (eg if flushed overnight with high-flow oxygen) and depend on the type and quantity of strong base in the adsorbent (in descending order of reactivity these are KOH > NaOH >> Ba[OH]2). Sevoflurane releases more heat than desflurane or isoflurane,208 and carbon monoxide production is highest with desflurane > enflurane > isoflurane > sevoflurane > halothane.209,210

Photochemical Breakdown of Anesthetic Waste Gases

Waste gases scavenged from anesthesia machines enter the atmosphere, where they are exposed to solar radiation and other gases.211,212 Ultraviolet light catalyzes the reaction between N2O and O2, producing the free radical nitric oxide, which in turn destroys atmospheric ozone.213 Waste N2O from medical uses is about 3% of the total emissions (the majority results from agriculture and combustion of fossil fuels). Halogenated volatile anesthetics act as greenhouse gases.214 Halogenated anesthetics also break down when exposed to ultraviolet light to form halogen free radicals, which deplete atmospheric ozone. Volatile anesthetic waste gases represent only a small portion of total atmospheric chlorofluorocarbons.

Biotransformation of Inhaled Anesthetics

Hepatic Drug Metabolism

In the liver, enzymes can transform volatile anesthetics by oxidation, reduction, and conjugation. These reactions convert hydrophobic substrates into more hydrophilic metabolites that are excreted via the kidneys. Of these, the most important pathways for volatile anesthetics are oxidative, and the enzymes responsible are various members of the large cytochrome P450 (CYP) family. Neonates lack some enzymes that are present in older humans, and diverse other factors, such as genetic variation, can alter individual metabolic activities. Intrinsic liver disease or hepatic congestion caused by heart failure may result in diminished enzymatic capacity, and intrahepatic blood flow shunting may reduce the efficiency of drug metabolism. CYP enzyme activities may also be inhibited by drugs such as cimetidine and amiodarone or enhanced by prolonged exposure to "inducers" such as phenobarbital, phenytoin, and a wide range of other compounds, including inhaled anesthetics.

Hepatitis Associated with Volatile Anesthetics

Among the currently available agents, halothane undergoes the most hepatic metabolism (20%-25%) and is associated most frequently with significant toxicity. Usually, less than 1% of halothane metabolism is reductive. Halothane's oxidative metabolism by CYP enzymes generates trifluoroacetic acid, bromide, and an intermediate metabolite, trifluoroacetic chloride, which can covalently modify proteins. Protein acetylation by these compounds primarily occurs within the liver, and modified proteins can act as neoantigens that stimulate the immune system to attack hepatocytes, resulting in fulminant hepatic necrosis. "Halothane hepatitis" occurs in 1 in 6000 to 1 in 35 000 adults after halothane anesthesia and is fatal in 50% to 75% of these cases.215 Genetic factors are likely involved, and females are affected about twice as frequently as males. Hepatic necrosis has been reported in many cases after multiple exposures to halothane, consistent with an amplified secondary immune response. Halothane hepatitis has also been reported in pediatric patients, but its incidence in children is 10 to 20 times lower than in adults.

Other inhaled anesthetics are metabolized by CYP enzymes to reactive acetyl intermediates that can also modify hepatic proteins. Rare cases of fulminant hepatic injury after administration of enflurane, isoflurane, and desflurane have been reported and the incidence for each anesthetic is related to the degree of their metabolism (2.5%, 0.2%, and 0.02%, respectively).216

Fluoride Nephrotoxicity

Metabolism of some inhaled anesthetics releases inorganic fluoride ions (F–), which can cause polyuric renal failure and increased mortality. Clinical findings include hypernatremia, hyperosmolarity, and increased BUN and creatinine. This problem is primarily associated with methoxyflurane, a very blood- and tissue-soluble anesthetic that is no longer in use. Metabolic release of fluoride is linked to methoxyflurane nephrotoxicity in a dose-related manner, and its causal relationship is shown in animal experiments. Renal injury after methoxyflurane is rare at fluoride blood levels below 50 μM; moderate injury is seen at 50 to 80 μM and severe injury at higher levels. In rats, clinical and pathologic changes similar to those in human fluoride nephrotoxicity can be produced by IV administration of fluoride at similar levels.

Other fluorinated anesthetics, particularly enflurane and sevoflurane, release detectable amounts of fluoride when metabolized in the liver. Nephrotoxicity, however, is very rarely associated with these drugs. Serum fluoride levels during and after enflurane anesthesia are usually below 50 μM, but those during and after sevoflurane anesthesia often peak above 50 μM. However, sevoflurane is much less soluble in blood and tissue than methoxyflurane. Sevoflurane is eliminated rapidly at the end of anesthesia, which halts fluoride production. In contrast, methoxyflurane is retained in tissues, and fluoride concentrations continue to increase after delivery is halted, reaching peak levels 2 to 3 days after anesthesia (for a review, see Anders217). Moreover, methoxyflurane is decomposed to fluoride within the kidneys more than sevoflurane, so intrarenal fluoride levels are higher than those detected in blood samples.218

Although the safety of inhaled anesthetics has improved greatly since the 19th century, all of the currently used agents are far from ideal because of their toxic effects on various physiologic systems (Table 38-9). The inert gas xenon was shown to produce anesthesia in 1951, and further study has revealed a pharmacokinetic and pharmacodynamic profile that approaches the ideal for an inhalational anesthetic.219,220 Xenon is an odorless, tasteless, nonflammable, and nonexplosive gas. Similar to other noble gases (helium, neon, argon, krypton, and radon), it is extremely chemically inert, undergoes no metabolic transformation, and has no direct negative environmental effects. Xenon does not cause airway irritation, and its blood–gas partition coefficient is 0.14, which makes inhalation induction significantly faster than with N2O or desflurane. Xenon, similar to N2O, produces minimal cardiovascular or respiratory depression. It produces no direct systemic organ toxicity, and there is evidence that xenon has neuroprotective effects,221 including the attenuation of cognitive dysfunction after cardiopulmonary bypass in an animal model.222 However, clinical studies have not demonstrated this benefit in patients.223,224 Because MAC for xenon is about 0.7 atm for patients near age 40 years and ED95 ≈ 1.3 × MAC ≈ 0.9 atm, its use as a sole anesthetic agent is likely limited to older patients who require lower concentrations. One of its few negative physiological effects is, similar to N2O, airspace expansion. Xenon has analgesic effects, which are likely attributable to inhibition of N-methyl-D-aspartate (NMDA)–sensitive glutamate receptors. Unlike ketamine, an IV anesthetic that also acts on NMDA receptors, xenon is not associated with emergence delirium.

The main barrier to routine use of xenon is the high cost of its fractional distillation from air, where its concentration is about 1 part per 12 million. Xenon cannot be synthesized, so these factors are unlikely to change unless new sources are discovered. Energy expended in collecting xenon also creates secondary negative environmental impacts. Methods for conserving (closed circuit delivery) and recycling xenon may make its clinical use economically possible in the future.

• Anders MW. Formation and toxicity of anesthetic degradation products. Annu Rev Pharmacol Toxicol. 2005;45:147-176.
• Avidan MS, Zhang L, Burnside BA, et al. Anesthesia awareness and the bispectral index. N Engl J Med. 2008;358:1097-1108.
• Baum JA. Low Flow Anaesthesia: The Theory and Practice of Low Flow, Minimal Flow and Closed System Anaesthesia. 2nd ed. Boston, MA: Butterworth-Heinemann; 2001.
• Campagna JA, Miller KW, Forman SA. Mechanisms of actions of inhaled anesthetics. N Engl J Med. 2003;348:2110-2124.
• Eger EI. Uptake of inhaled anesthetics: the alveolar to inspired anesthetic difference. In: Eger EI, ed. Anesthetic Uptake and Action. Baltimore, MD: Williams & Wilkins; 1974:77-96.
• Eger EI 2nd. Age, minimum alveolar anesthetic concentration, and minimum alveolar anesthetic concentration-awake. Anesth Analg. 2001;93:947-953.
• Kenna JG, Jones RM. The organ toxicity of inhaled anesthetics. Anesth Analg. 1995;81(Suppl):S51-S66.
• Myles PS, Leslie K, McNeil J, et al. Bispectral index monitoring to prevent awareness during anaesthesia: the B-Aware randomised controlled trial. Lancet. 2004;363:1757-1763.
• Sebel PS, Bowdle TA, Ghoneim MM, et al. The incidence of awareness during anesthesia: a multicenter United States study. Anesth Analg. 2004;99:833-839; table of contents.
• Stoelting RK, Eger EI 2nd. An additional explanation for the second gas effect: a concentrating effect. Anesthesiology. 1969;30:273-277.