Chapter 43. Total Intravenous Anesthesia

1. The availability of drugs with short blood–brain equilibration times (especially those with an ester linkage) enables the clinician to use intravenous anesthetics and analgesics where controllability is easy and recovery rapid.

2. Total intravenous anesthesia (TIVA) offers some important advantages over inhalation anesthetics, including rapid recovery with minimal hangover and a low incidence of nausea and vomiting. TIVA may be the technique of choice for some operations.

3. Effective delivery of TIVA requires the clinician to have a good knowledge and understanding of pharmacokinetics, pharmacodynamics, and pharmacokinetic–pharmacodynamic (PK-PD) modeling.

4. Important drug characteristics include induction dose, rate of administration, and ke0 (rate constant for the elimination of drug from the effect compartment; drugs with small ke0 values take longer to equilibrate between the blood and the effect compartment or biophase). Thus for a rapid sequence induction, drugs with a large ke0 are preferable (viz., propofol, thiopental, remifentanil, and alfentanil compared with midazolam, ketamine, and fentanyl).

5. Drug interactions are important in TIVA. Excepting ketamine, opiates and hypnotics potentiate each other and can result in synergistic cardiovascular and respiratory depression. This means the doses of each can be reduced. There is a ceiling to the potentiating effect of the opiates, beyond which there is an increased incidence of adverse side effects and often delayed onset of spontaneous ventilation at the end of surgery.

6. Ketamine is the only hypnotic drug with established analgesic properties; there are few data suggesting that the S (+) isomer has major advantages over the racemic mixture.

7. If the patient shows signs of response during TIVA, additional IV supplementation can be achieved by bolus doses or by increasing the infusion rate, or the targeted concentration if using target-controlled infusion (TCI).

8. The elimination half-life and systemic clearance are not useful in TIVA for determining the offset of IV drugs. The context-sensitive half-time (CSHT) is more relevant, and modeled data and dynamic measurements correlate well.

9. Context-sensitive decrement times are useful indices of recovery from anesthesia. Recovery is influenced by other factors apart from the CSHT, including patient characteristics such as age, sex, body habitus, coadministered therapies, and disease states.

10. Knowledge of the context-sensitive decrement times for opiates and hypnotics allows appropriate choice of drugs for the maintenance of anesthesia. CSHT also helps determine which infusions should be terminated at the end of anesthesia, and which should be terminated some time beforehand.

In 2011 we are still without any one drug that can provide all the requirements of anesthesia (ie, unconsciousness, analgesia, amnesia, and muscle relaxation). Consequently, administration of several different agents is needed to produce the desired end result. The use of intravenous agents to achieve these goals began with the introduction of the rapidly acting barbiturates in 1934. Despite the disastrous consequences after the use of thiopental at Pearl Harbor in 1941, intravenous anesthesia is now well established as an appropriate alternative to the traditional approach of volatile anesthetics alone; indeed, sometimes it is the preferred alternative. Further, intravenous anesthesia is a fundamental component of "balanced anesthesia," in which volatile and intravenous anesthetics are combined to produce the anesthetic state.

The kinetics of the early barbiturates did not render the drugs ideal for the maintenance of anesthesia; the provision of analgesia could not be achieved by the barbiturates alone, and the addition of either meperidine or morphine (both drugs having slow blood–brain equilibration) led to overdosing and hence to poor clinical conditions, especially in the spontaneously breathing patient. The introduction of modern volatile agents, starting with halothane in 1956, with their easy titratability, encouraged the anesthetist to turn away from the intravenous agents for the maintenance of anesthesia.

The development of and present interest in total intravenous anesthesia (TIVA) owes much to a number of important studies. When considering the role of hypnotic agents, the contribution of Savege et al,1 who, in 1975, used the steroid agent Althesin (alphaxalone and alphadolone acetate) (Glaxo, London, UK) with meperidine to supplement oxygen-enriched air in the spontaneously breathing patient, is of great importance. Subsequent drug developments included the use of infusions of thiopental, methohexital, the carboxylated imidazole etomidate, propofol, midazolam, and ketamine. With the exception of ketamine, none of these agents probably provided analgesia (although analgesic effects of alphadolone have been described by Goodchild et al2 when administered by the enteral route).

Another approach to the provision of intravenous anesthesia was studied by Lowenstein et al3 in 1969 using morphine as a sole agent to achieve anesthesia in patients undergoing cardiac surgery. This method of anesthesia (more popular in the United States and Europe than in the United Kingdom) was not reliably effective. It was associated with a high incidence of episodes of awareness during anesthesia, as well as intraoperative episodes of hypotension caused by histamine release, resulting in increased intraoperative and postoperative fluid and blood requirements. A reduction in the incidence of some of these side effects was seen when morphine was replaced with fentanyl or one of its congeners, but at the cost of increased truncal rigidity.

The selective and highly potent μ-receptor opiates (fentanyl, alfentanil, sufentanil, and more recently, remifentanil) have the advantage of less cardiovascular depression, as well as the ability to obtund the hemodynamic responses to laryngoscopy and intubation. However this apparent stability is less evident during major cardiovascular surgery, particularly during the period of sternotomy and aortic root dissection in patients undergoing cardiopulmonary bypass, where there are frequent episodes of significant hypertension and tachycardia that are not always remediable by increased doses of the opiates. Thus it became clear that use of high doses of opiates alone was inappropriate to provide anesthesia. Hence attempts to control the hemodynamic effects of surgical stimulation and to reduce the incidence of awareness have used either low inspired concentrations of volatile agents or infusions of a sedative-hypnotic agent as supplements to opiates.

Use of TIVA techniques are not without difficulties in these patients, as some combinations of opiates, such as fentanyl and diazepam, can result in significant reductions in both cardiac output and systemic vascular resistance. However, the effectiveness of the benzodiazepines in reducing the circulating levels of catecholamines may be important, and greater hemodynamic stability has been found in cardiac surgical patients in whom the combination of sufentanil with ketamine or midazolam has been employed. In noncardiac surgery, opiates remain popular as the basis of most neurosurgical anesthetic techniques, as they do not alter the carbon dioxide reactivity of the cerebral blood vessels. The kinetics and dynamics of remifentanil (and to less extent alfentanil) allow the anesthesiologist to titrate dose requirements more closely for a given surgical procedure during abdominal and other major body surface surgical procedures.

Over the past 3 decades, increased interest in intravenous anesthesia has been prompted by the availability of newer and more appropriate rapid-onset and short-acting IV sedative-hypnotics (eg, midazolam, propofol, etomidate), analgesics (alfentanil, sufentanil, remifentanil), and muscle relaxants (atracurium, vecuronium, mivacurium, and rocuronium). All these factors have refocused the anesthetist's attention on the total provision of anesthesia by the intravenous route. Furthermore, there are also a number of clinical situations in which intravenous anesthesia techniques are advantageous. These include

• Provision of sedation to supplement local or regional anesthetic techniques
• Provision of general anesthesia with accompanying minimal cardiovascular depression
• Use in anesthesia for office or ambulatory surgery, where the speed and completeness of recovery are important
• Use in situations in which volatile-based anesthetics may be difficult to administer because of the unavailability of resources, such as nitrous oxide; at sites of military or nonmilitary trauma; and for anesthesia at increased ambient pressure
• Use in circumstances in which nitrous oxide may either be undesirable (eg, because of the need for high inspired oxygen concentrations) or contraindicated (eg, for 1-lung anesthesia; for middle ear surgery, some neurosurgical operations, and prolonged abdominal surgery where nitrous oxide can cause closed-space effects as a result of increased volume and pressure; for relief of cardiac tamponade; for airway endoscopies [bronchoscopy and laryngoscopy]; and for bronchotracheal surgery), or where volatile agents may best be avoided (Table 43-1)
• Prevention of awareness during cardiopulmonary bypass and cerebral protection in patients at risk of episodes of brain ischemia
• Use in malignant hyperthermia-susceptible patients

However, the general application of TIVA techniques is not without important considerations and questions. These include the possibility and detection of awareness, the possibility of postoperative respiratory depression because of the persistent effects of concurrently administered analgesic agents, the requirement of a separate dedicated IV access site, appropriate infusion pumps, and the observation that some researchers consider depth of anesthesia as uncontrollable as that with the volatile agents.

What Are the Ideal Drug Properties for and Advantages of TIVA?

Table 43-2 lists the ideal properties of a hypnotic drug for continuous infusion anesthesia. Although none of the present hypnotic or analgesic agents fulfills all of these criteria, some are more suitable than others under differing circumstances. Table 43-3 lists some of the pharmacologic advantages of TIVA over inhalational agents.

Dosing Strategies for TIVA

The administration of hypnotic, analgesic, and neuromuscular blocking components of TIVA can be provided in a number of different ways:

• Intermittent bolus doses of the drugs
• Continuous infusions using syringe infusion pumps or similar delivery systems
• Through the introduction (in most parts of the world, although not presently in the United States) of target-controlled infusion systems (TCI)

When given by intermittent bolus doses, there will be fluctuating drug concentrations that are associated with accompanying changes in the depth of anesthesia. Continuous infusions of IV hypnotics, opiates, and muscle relaxants can reduce these swings in drug concentrations, as well as minimize any relative under- or overdosing during drug administration. This offers a number of advantageous features to the anesthetized patient, including greater hemodynamic stability, fewer episodes of hemodynamic breakthrough and other signs of patient responsiveness, reduced need for supplemental anesthetics or vasoactive drugs, more rapid awakening and decreased incidence of requirements for naloxone or need for postoperative ventilatory support, decreased incidence of side effects, and a 25% to 30% lower total drug dose compared with bolus dosing.5-9 Use of continuous drug infusions may also lead to decreased drug costs.

However, because modern IV agents exert profound pharmacologic effects and have short durations of action, frequent adjustments to drug dosage are necessary in order to match anesthetic effect to the varying noxious stimuli being perceived at different times during surgery, as well as the different noxious intensities of different surgical operations. This requires either frequent additional bolus dosing or the calculation of new infusion rates to achieve different drug targets.

Early studies using IV infusion anesthesia focused on providing fixed rate infusions that were adequate for 50% or more of a given population.6,10 With the passage of time, these have been viewed as being less appropriate and flexible in output because they failed to allow easy titration of the achieved drug concentration to clinical effect in accordance with an individual patient's requirements. Indeed there is the probability that many patients received an overdosage because of the desire to design regimens that were "all-embracing" to all people! Nevertheless, the work of Prys-Roberts11 and Sear12,13 confirmed the usefulness of this approach to the provision of anesthesia with minimal cardiorespiratory depression and no reported cases of awareness.

With the hypnotic drugs currently available for continuous infusion, several manual protocols for anesthesia delivery have been developed. For example, there is the regimen for propofol described by Roberts et al14 based on an infusion protocol to maintain a plasma propofol concentration of approximately 3 μg/mL. This was achieved by an initial 1 mg/kg induction dose, then 10 mg/kg/h for 10 minutes, 8 mg/kg/h for 10 minutes, and maintained at 6 mg/kg/h. Because this scheme is designed to achieve a fixed plasma propofol concentration, there is often a need for supplementation with additional propofol boluses and/or opiates during abdominal or other major operations. Similar manual regimens have been described for midazolam infusion. Persson et al15,16 used an induction dose of 0.25 mg/kg, followed by a fast infusion of 0.65 mg/kg/h for 15 minutes and maintained with 0.13 mg/kg/h. This provided a hypnotic plasma concentration of 300 to 400 ng/mL. In a separate study, Thiel et al17 described an infusion of midazolam (0.1-0.15 mg/kg then 0.02 mg/kg/h) supplemented with an opiate infusion of sufentanil for cardiac surgery.

Pharmacokinetic Principles of TIVA

Simplistically, the design of infusion protocols for TIVA is based on 2 important equations that define the loading dose and the maintenance infusion rate.

where Vd = initial apparent volume of distribution, Cp = the desired plasma drug concentration, and Cl = the systemic clearance of the drug.

However, these calculations of drug requirements to reach a given target concentration are flawed for several reasons. First, the plasma is not the site of action of IV drugs; the site where the drugs produce their effects is in the brain and is termed the biophase. To reach the biophase, drug redistributes from the blood to the brain. At the same time, drug is also being redistributed to other tissues of the body. Hence the loading dose necessary to produce a desired pharmacologic effect cannot usually be calculated based on the initial volume of drug distribution (which is primarily the blood volume), but should use the apparent volume of distribution into which the drug has distributed once it has equilibrated with the biophase (ie, this will, of necessity, be a larger volume). (See Chapter 40 for a detailed discussion of pharmacokinetic [PK] and pharmacodynamic [PD] principles.)

When a drug is given by rapid infusion, there may be a simultaneous pharmacologic effect; however, the measure of drug effect (be it the electroencephalogram [EEG] spectral edge frequency, minute ventilation volume, or change in blood pressure or heart rate) does not always parallel the rapid increase and decrease in the plasma drug concentration. This implies "hysteresis" in the relationship between drug concentration and effect. From studies that provide continuous measurement of plasma drug concentration and effect, it is possible to relate the plasma drug concentration to the effect produced in the biophase18 and, in turn, to calculate the volume of distribution of this effect compartment.

Further, by complex mathematical manipulation, the hysteresis loop can be "collapsed" to derive a linear relationship between concentration and effect. The value that causes the hysteresis loop to collapse represents the rate of equilibration of the drug concentration between the plasma and the biophase. This is termed ke0 or the blood–brain equilibration rate constant. Table 43-4 shows the values of ke0, t½ke0, time to peak effect, and the apparent volume of the effect compartment.

Time to Peak Effect

Plasma drug concentration targeting is associated with a temporal delay between the equilibration of concentrations in the plasma and at the effect site. The rate of equilibration depends on 2 factors: the rate of drug delivery to the effect site (which is influenced by cardiac output and cerebral blood flow) and the pharmacologic properties of the drug with respect to passage across the blood–brain barrier.

Although the t½ke0 is the time for blood–brain equilibration, this can only be defined if both the kinetic and dynamic properties of the drug are measured simultaneously. Because of the interindividual variability of both of these properties, it is recommended that if PK and PD data are not available from the same subjects that the time to peak effect (TTPE), a model-independent parameter, be used to estimate the average ke0 for a given kinetic model and given patient group.19 In general, the TTPE for a drug is independent of drug dosage.

For simple infusion models (eg, the Marsh kinetic model used in the Diprifusor20 [Zeneca Pharmaceuticals, Macclesfield, United Kingdom]), the same bolus dose will have the same predicted drug concentration profile, the same peak concentration, the same TTPF, and therefore the same ke0. For more complex models (as those described by Schnider et al 21 and Minto et al,22 in which there are included a number of covariates such as age, height, weight, and lean body mass), there will be different peak concentrations, different drug time courses, and different TTPF for patients with different characteristics (ie, each patient will have a unique ke0).

For optimal dosing strategies, clinicians need to know the TTPF when giving IV drugs during both the induction and maintenance phases of anesthesia (Fig. 43-1). For example, during a rapid sequence induction, the anesthesiologist should use drugs with short times to peak effect. The combination of thiopental or etomidate and alfentanil or remifentanil is probably more appropriate than a combination of propofol and fentanyl. With the latter combination, the effect of the fentanyl (if given immediately before the hypnotic agent) would not be maximal by the time of intubation, leading to an initial hypertensive response to the noxious stimulus of laryngoscopy. This might then be followed by a hypotensive response when the peak effect of the fentanyl is achieved and airway stimulation is minimal. In addition, when giving drugs by bolus dosing, the interval between doses needs to be of a sufficient duration to allow the peak effect of 1 bolus to be observed before the next dose of the drug is administered.

For most of the drugs used as the hypnotic and analgesic components of TIVA, plasma (blood) concentration–effect relationships have now been determined. From these, and using appropriate kinetic data sets, the clinician can derive manually controlled dosing schedules (Tables 43-5 and 43-6). However, these protocols do not really take account of time delays in the concentration–effect relationship. Hence many anesthesiologists view TIVA using manually controlled infusion systems and devices as a more complex strategy for maintaining anesthesia than using a vaporizer in the case of volatile agents!

Computerized Drug Administration: Target-Controlled Infusions

One of the most significant and recent advances in the delivery of infusion anesthesia over the last decade is the introduction of TCI systems that allow the anesthesiologist to set the target concentration of anesthetic. This target can be defined as either the blood or the effect-site (ie, brain) concentration. In many respects, these systems function in a similar manner to the setting of the inspired concentration of the volatile agents using a vaporizer. To achieve a given target concentration requires a greater initial infusion rate followed by a declining maintenance rate over time.

A number of publications describe a theoretical approach to attaining and maintaining steady-state drug concentrations,23-26 with different algorithms used for targeting blood concentrations5,27-29 and incorporating effect-site concentrations into model-driven automated delivery systems.30,31 Several different systems for computerized drug delivery have been described. For example, Schwilden et al32 formulated a system to maintain anesthesia using the hypnotic agent etomidate. Other systems include the computer-assisted continuous infusion (CACI) system for fentanyl5; the target-controlled infusion-based anesthesia (TIAC) system, also for fentanyl6; and the various systems for propofol.33

A TCI system consists of an infusion pump linked to a microprocessor containing the disposition pharmacokinetic parameters for the drug in question. The equations describing disposition are based on the bolus-elimination-transfer (BET) scheme of Schuttler et al.34 TCI combines sequential BET schemes for maintaining or increasing the target concentration and periods of discontinuation of the infusion when the target concentration has to decrease. Infusion halts are mandatory if a new lower drug concentration is to be achieved as fast as possible. The first commercially available TCI system is the Diprifusor (Zeneca Pharmaceuticals),35 which was introduced in 1996 and uses prefilled glass syringes of propofol tagged with a unique metallic strip. More recently, open TCI systems that allow use of the various available generic formulations of propofol have been developed. Such systems include the Asena PK system (Alaris Medical Systems, San Diego, CA), the Base Primea system (Fresenius Kabi, Bad Homburg, Germany), and the FM System (B Braun, Melsungen, Germany).

At present, no drugs apart from propofol are licensed for use for TCI; in addition, approval for TCI is still awaited in the United States. However, research or developmental computer-driven delivery systems have been described for both sufentanil and remifentanil.36–41 Sufentanil has been administered by TCI in combination with isoflurane, propofol, and midazolam, with targeted concentrations, in most studies, of sufentanil between 0.2 and 1.5 ng/mL for noncardiac surgery and 3 and 10 ng/mL for cardiac surgery. Most studies report good-quality anesthesia with hemodynamic stability and satisfactory times to return of spontaneous ventilation and extubation. Remifentanil by TCI has been assessed in combination with both IV hypnotics and volatile agents,19,42-44 the most appropriate kinetic model for drug delivery being that of Minto et al.22 For analgesia, concentrations between 1 and 1.5 ng/mL are needed. In combination with propofol for noncardiac surgery, remifentanil concentrations in the range of 4 to 8 ng/mL are appropriate. However, these data need confirmation in larger clinical trials.

A TCI drug delivery system has 4 key components:

1. A calculator (computer or microprocessor) and a software program, either separate with cables linking to a syringe pump, or an "all-in-one" system (as is the case with the Diprifusor)

2. Data sets of pharmacokinetic parameters for each drug to be infused

3. An infusion device

4. A user interface

For TCI delivery, the anesthesiologist needs to enter the patient's age and weight into the device and then select the desired target drug concentration. On pressing "start," the TCI system delivers precisely the amount of drug to achieve the target drug concentration and then continues to infuse drug at an appropriate rate to maintain that concentration.

Unlike using a manual fixed-infusion system, the anesthesiologist can easily either increase or decrease the target drug concentration at any time, thereby facilitating the titration of the drug concentration according to the patient's response to differing noxious stimuli.

One fundamental principle underlies the TCI pump—namely, the relationship between the infusion rate and achieved drug concentration:

If a higher target concentration is selected, the system automatically delivers a further small bolus of drug and then infuses at a faster rate to achieve and maintain the higher drug concentration. When a lower target is chosen, the device discontinues the infusion of drug until the system predicts that the patient's drug concentration has decreased to the new value (by a combination of drug redistribution and clearance). The system then restarts the infusion at the appropriate rate for the new target concentration. With a TCI system, the anesthesiologist does not control the device's infusion rates and the size of bolus doses; rather, the anesthesiologist titrates the depth of anesthesia by selecting target drug concentrations appropriate to the patient's needs. A number of different algorithms are used in the different systems to calculate the required infusion rate.28,31,33,45

Kinetic Model Selection

TCI systems tend to incorporate population-averaged pharmacokinetic parameters to drive the infusion pump. This can lead to considerable variability in the observed drug concentration compared with the targeted concentration. The mean variation between these 2 concentrations (targeted and observed) is usually of the order of 20% to 30%, with a maximum of 50% to 60% considered clinically acceptable. It seems likely that greater accuracy is not going to be achievable. The sources of this variability include the differences in the parameter estimates of the polyexponential equation that is used to describe the concentration–time relationship. At the present time, there are several different commercial systems on the market for the delivery of propofol and the opiates sufentanil and remifentanil. Two types of systems exist presently—those incorporating the Diprifusor module, using tagged Diprivan (propofol; AstraZeneca, London, United Kingdom) prefilled syringes with the fingergrip tag and the original Diprifusor kinetics and "open" systems that do not require the tagged syringes and use different pharmacokinetic data sets for the delivery of propofol, remifentanil, and sufentanil (Table 43-7). For further discussion of TCI open systems, the reader is referred to the Open TCI Initiative (www.opentci.org).

Effect-Site Targeting

Initial drug-targeting technology was based on the plasma concentration, but as this is not the site of anesthetic action, it is more logical to target the effect site or biophase. This has been described by several authors and reviewed by Jacobs and Williams.30 Effect-site control should allow for a faster induction time and during maintenance a more rapid increase in the depth of anesthesia. However, there are some limitations to effect-site compartment targeting, including that of overshoot of the arterial drug concentration with associated adverse consequences. Effect-site TCI has been described by a number of different research groups.8,30,46

Wakeling et al47 have compared the efficacy of plasma and effect-site targeting. Twenty American Society of Anesthesiologists (ASA) classification PS-I or PS-II patients were randomized to receive TCI propofol targeted at one or the other site. In this study, the delivery system was modeled according to the kinetic data set of Gepts et al48 and a ke0 value of 0.63 min–1. At a target concentration of 5.4 μg/mL (the predicted effect-site concentration associated with loss of consciousness in 95% of subjects),49,50 Wakeling found a median time to loss of consciousness of 3.02 minutes in the plasma-targeted group, but only 1.23 minutes in the effect-site targeted group. However, loss of consciousness occurred at the same predicted effect-site concentration in both groups.47 Other studies of effect-site drug control have found similar results.46,51,52

Although an advantage of these systems is that drug delivery is optimized to a given effect-site concentration, the speed of achieving this concentration varies with the ke0 value incorporated into the different delivery systems. At present, several different models with different ke0 values are in use for the delivery of propofol. If the Diprifusor is used with the Marsh model kinetics, the blood–brain equilibration half-life (t½ke0) is 2.6 minutes (ke0 0.26 min−1).53 However, modifications of the Diprifusor algorithm in the 2 new open-TCI systems (the Base Primea and Asena PK systems) with a longer ke0 of 1.2 min−1 gives a t½ke0 of 34 seconds. When the slower ke0 is used, any increase in the target effect-site concentration requires higher blood drug concentrations to generate a bigger gradient to drive drug to the effect site.

A second effect-site model, based on the kinetics of Schnider et al,21 incorporates additional details into the algorithms (including patient age, height, weight, and calculated lean body mass). In this system, the ke0 varies with the age of the patient. Although most studies suggest that effect-site targeting provides faster control, it is only true if there is an accurate value for ke0! The recent studies of Kazama et al54 show that ke0 values have a wide interindividual variability with regard to age and end point assessed. This may be a result of a number of factors, including poor delineation of early distribution kinetics and failure to accurately define the EEG hysteresis effect; this will clearly limit the accuracy of any effect-site TCI delivery system in a given individual. Use of effect-site TCI will, of necessity, lead to an overshoot of the plasma drug concentration, which can result in hemodynamic depression. Although this overshoot has little effect in patients younger than age 60,46 further study is needed in the elderly.

As a further development in effect-site targeting, Van Poucke et al55 have described modified delivery algorithms that target the effect-site but limit the peak plasma concentration. This should have the advantage of reducing the acute hemodynamic effect of intravenous anesthetic drugs on the heart and circulation. Simulations suggest that this approach will reduce the plasma overshoot by approximately 60%, with an accompanying 20% delay in the time to peak effect and hence the onset of drug effect. In vivo studies are clearly needed to confirm the efficacy of this method of TCI delivery.

Several factors affect the speed of induction by TIVA, including the size of the induction dose and important drug interactions.

Factors Determining the Induction Dose of Hypnotic Agents

The onset of anesthesia requires the brain drug concentration to reach a given level; this can be achieved either slowly or rapidly. Rapid achievement is usually accompanied by significant adverse effects such as hypotension, bradycardia, and respiratory depression. The greater the gradient, the greater the time needed to induce anesthesia. Conversely, any marked overshoot in the effect-site concentration leads to a greater incidence and severity of the adverse effects. The transfer of drug from the blood to the effect site is governed by simple diffusion. The time needed to achieve this transfer varies with the concentration gradient and the ke0.

The rate of infusion of the induction dose is another determinant governing the size of the induction dose; an infusion rate designed to only just achieve a desired effect-site concentration causes a loss of consciousness but at the cost of a slow onset. The loss of consciousness is only transient, and is maintained only for the duration that the targeted effect-site concentration is maintained. A faster rate of infusion provides a more rapid onset of anesthesia and a longer duration of loss of consciousness, but again involves administration of a larger induction dose and hence a greater likelihood of adverse effects.

Variations in induction dose requirements also occur because of interindividual pharmacokinetic differences and because of pharmacodynamic differences that are the result of age, sex, cardiac output, smoking, concomitant medications, and the presence of coexisting disease states.56

Drug Interactions

Because no single IV drug (with the possible exception of ketamine) can provide hypnosis, amnesia, and analgesia, total intravenous anesthesia will be, by definition, provided by a combination of drugs.57 However, in contrast to the simply additive effects of the different gaseous and volatile anesthetics (see Chapter 38), IV anesthetics may interact in either additive or synergistic manners (see Chapters 41 and 42). In the case of most IV agents, this interaction is synergistic. One example used to good effect is the combination of a hypnotic and benzodiazepine or other sedative drug (eg, an α2-agonist) for induction of anesthesia (the technique of coinduction).57 One major advantage of this technique is the provision of an adequate depth of anesthesia for a noxious stimuli such as laryngoscopy and intubation without any associated significant cardiovascular depression.

Also clinically useful is pretreatment with a rapid-acting opiate (eg, fentanyl, sufentanil, alfentanil, or remifentanil), which will reduce the amount of hypnotic agent needed for loss of consciousness and provide analgesia to reduce the adrenergic stimulus caused by the instrumentation of the airway (whether by laryngoscopy and intubation or insertion of the laryngeal mask airway). However, the combination of an opiate with an IV hypnotic agent can cause both positive and negative side effects (namely, cardiovascular stability or profound cardiovascular and respiratory depression). Respiratory depression is of little importance in the mechanically ventilated patient, but is a significant and unwanted effect in the spontaneously breathing individual.

Although the combination of benzodiazepines with nitrous oxide produces minimal cardiovascular effects for induction of anesthesia, the combination of benzodiazepines with opiates results in a synergistic depression of most hemodynamic parameters. Because opiates alone are not complete anesthetics, they cannot be used reliably to induce anesthesia and amnesia in the absence of supplements. In clinical practice, a second anesthetic drug (eg, an IV hypnotic) should always be given to ensure loss of consciousness. Thus the main functions of opiates are to provide analgesia and potentiate the effects of IV anesthetic agents, both at induction and during the maintenance phase.

Termination of IV Anesthetic Action

The duration of drug effect of a bolus dose of IV drugs is terminated predominantly by redistribution of drug from the blood and brain to the lean tissues. This redistribution occurs by a series of different compartmental clearances, and to a lesser extent, metabolic clearance, the sum of these depending on cardiac output in the case of lipid-soluble hypnotic agents. However, if the dose of the induction agent causes a significant reduction in cardiac output, the offset of the drug effect is delayed.

(Note: Detailed discussions of the following drugs appear in Chapters 41 and 42.)

Thiopental

This thiobarbiturate was the most popular of all IV anesthetic induction agents, although propofol may now be more appropriate in circumstances where rapid recovery from anesthesia is a prerequisite (eg, in ambulatory surgical practice). Thiopental has a number of drawbacks as it is highly alkaline in solution and therefore extremely irritating if injected extravascularly or intra-arterially, and it has a low therapeutic index of about 4. When given IV, the barbiturates achieve onset of hypnosis within 1 arm–brain circulation time. Because of their high lipid solubility and low percentage of ionization at physiologic pH, there is rapid uptake into the brain. The maximum effect of thiopental is seen within approximately 1 minute of a single induction dose of 3 to 6 mg/kg. There then follows a similarly rapid decline in the brain concentration as a consequence of redistribution into the "lean" body tissues, making the duration of effect of an induction dose about 5 to 10 minutes. However, at the time of awakening, only 18% of the injected thiopental dose will have undergone metabolism (compared with approximately 38% of a dose of methohexital and nearly 70% of propofol).

Population kinetic data for thiopental have been described by Stanski and Maitre, with a mean clearance value of 0.0031 L/min/kg and apparent volume of distribution at steady state of 2.73 L/kg.58 A number of different factors influence thiopental elimination. Although age influenced the kinetics of thiopental, it had no effect on brain responsiveness or dynamics when the spectral edge frequency was used as the measure of drug effect. The induction dose requirements for thiopental anesthesia vary with patient age and weight, and, most importantly, cardiac output. Lower doses are indicated in the premedicated patient, in patients with severe anemia or burns, in malnourished patients, in patients with uremia or liver failure, and in the hypovolemic individual regardless of cause. Both circulatory failure and hypothermia slow the circulation time, prolonging the induction period for thiopental. Moreover, the total dose needed in such patients is also reduced.

The blood–brain equilibration rate constant for thiopental is rapid (0.46-0.58 min–1). The effect-compartment concentration associated with dropping a loaded syringe is approximately 17 μg/mL, and the duration of effect of an induction dose of 320 mg is approximately 4 minutes. Induction with thiopental causes a decrease in cardiac output, and a compensatory 10% to 15% increase in heart rate. Thiopental also causes venodilation, but has no effect on systemic vascular resistance.59 Another feature of thiopental induction is the decrease in respiratory responsiveness to hypoxia and hypercapnia. Opiates potentiate these effects. When anesthesia is induced with the combination of thiopental and midazolam, the 2 hypnotics act synergistically; when thiopental is given with an opiate such as alfentanil, there is greater potentiation of the antinociceptive response than hypnosis.60

Thiamylal

Thiamylal (another thiobarbiturate used mainly in the United States and Japan) has pharmacologic characteristics similar to thiopental with regard to potency, incidence of laryngospasm and respiratory depression, cardiac toxicity, and recovery time. Like thiopental, it is formulated as a racemic mixture, with the potency of the S enantiomer about twice that of the R+ form. There are also enantiomeric kinetic differences. The plasma protein binding of the R+ isomer is 82.5%, and of the S– isomer 88.3%. The elimination half-life of the R+ enantiomer is 20.2 hours, the apparent volume of distribution is 3.66 L/kg, and clearance is 0.27 L/kg/h. The corresponding values for the S– compound are 24.1 hours, 2.60 L/kg, and 0.15 L/kg/h, respectively.

Methohexital

This is an oxybarbiturate with a faster recovery profile than thiopental because of a greater systemic clearance and a shorter elimination half-life. Compared with thiopental, methohexital has a more appropriate kinetic profile for both induction and maintenance of IV anesthesia (elimination half-life of 420-460 minutes, clearance of 700-800 mL/min). Because its main metabolite, 4-OH methohexital, has no pharmacologic activity (unlike thiopental's major metabolite, pentobarbital), methohexital may be usefully given by continuous infusion for maintaining anesthesia or sedation (see Drugs for Maintenance of TIVA).

Induction of anesthesia with methohexital requires doses of 1 to 2 mg/kg, giving it a potency of 2.7 times that of thiopental. The side-effect profile of methohexital is more significant than that of thiopental and includes increased incidences of pain on injection, a tendency to venous thrombophlebitis, and exaggerated involuntary movements, especially in the unpremedicated patient. Inadequate induction doses can also cause excitatory phenomena because the inhibitory areas of the brain are thought to be depressed at lower drug concentrations.

Etomidate

Etomidate is a carboxylated imidazole compound formulated as the R+ enantiomer. The approximate potency ratio of the enantiomers is R+:S– = 1:10.61 As an induction agent, etomidate has important advantages over the barbiturates, showing many ideal properties for an IV agent (cardiostability; reduction in cerebral blood flow, cerebral metabolic rate, and intracranial pressure; no release of histamine and low rate of allergic reactions; only transient and minimal respiratory depression and no inhibition of the hypoxic pulmonary vasoconstrictor reflex) and offers hemodynamic advantages during induction of anesthesia in patients with poor cardiac reserv, and hypovolemia. The drug has one of the widest margins of safety in animals of all the present hypnotic agents, with a high therapeutic index of 26.4.

Because it is unstable in water, etomidate is presently solubilized either in 33% propylene glycol or as an emulsion. The drug has a pH of 8.1 and pKa of 4.2. It is a base, and approximately 99% of the drug is un-ionized. Plasma protein binding is approximately 75% (mainly to albumen). Metabolism occurs predominantly in the liver and plasma by esterase hydrolysis, and hence etomidate shows the expected high systemic clearance.

Induction doses of 0.2 to 0.4 mg/kg provide hypnosis for 5 to 15 minutes, with only minor alterations in cardiovascular parameters in healthy patients and in those with valvular or ischemic heart disease. Little is known about the interaction of etomidate and opiates or other hypnotics for loss of consciousness. Etomidate alone does not obtund the sympathetic responses to laryngoscopy and intubation,62 and for a smooth hemodynamic profile, etomidate is best combined with an opiate or benzodiazepine.

However, its use as an induction agent is also associated with a number of minor disadvantages, with significant incidences of pain on injection, thrombophlebitis, myoclonia, and a high incidence of postoperative nausea and vomiting. When given either as a single induction dose or by infusion, etomidate suppresses adrenal steroidogenesis,63-65 and low plasma cortisol levels are found after use of the agent. Does this matter? In patients receiving etomidate by infusion for intensive therapy unit sedation, Ledingham and Watt66 found that its use was associated with increased mortality. However these findings were not based on a double-blind, randomized, controlled trial. There are no data to support this association when etomidate was administered by continuous infusion to anesthetized surgical patients.

The role of etomidate was further confused by an unreasoned and unsupported editorial in Anaesthesia expressing the view that the drug had no place in current anesthetic practice.67 This view is not shared by the present author in the absence of properly conducted clinical outcome trials in nonseptic patients. But what evidence is there? In 2010, Ray et al68 retrospectively reported on 176 patients admitted to an intensive care unit (ICU) after an emergency laparotomy. Fifty-two of the patients had anesthesia induced with etomidate. There was no association between use of etomidate and in-hospital mortality, and the risk of developing hypotension after the induction of anesthesia or the need for vasopressors to treat hypotension was least following etomidate. However, a second study (in the form of a randomized controlled trial of single doses of etomidate or ketamine in critically ill patients requiring emergency intubation) reported a greater percentage of patients with evidence of adrenal insufficiency in the etomidate group (odds ratio 6.7, 95% CI, 3.5-12.7).69

Use of etomidate for induction of anesthesia in the critically ill septic patient is probably contraindicated (despite the drug's minimal cardiovascular effects) because of data taken from a subpopulation within the CORTICUS trial that suggest an increased mortality.70,71

Alternate Formulations of Etomidate

Attempts at improving the side-effect profile of etomidate have focused primarily on the solvent. Development of an emulsion formulation does not change the drug's dynamic properties, but is associated with lower incidences of pain on injection, myoclonus, and local thrombophlebitis.72,73 Another advantage of the emulsion formulation is its lower osmolality and higher pH (400 mOsm/kg and pH 7.6 compared with 4965 mOsm/kg and pH 5.1 for the propylene glycol formation), which means less red cell hemolysis.74,75 A second reformulation of etomidate has used 2-hydroxypropyl-β-cyclodextrin as the solvent.76 Again, there was a lower incidence of myoclonia (17% vs 92%) and pain (8% vs 58%), and thrombophlebitis (0% vs 42%) and no hemolysis. None of these formulations appears to show alterations in the kinetics or dynamics of etomidate.

Although it is widely believed that use of etomidate is associated with an increased incidence of postoperative nausea and vomiting, this was not supported by a comparison in which etomidate-lipuro and propofol were used for induction of anesthesia to supplement isoflurane/fentanyl in air in patients undergoing orthopedic procedures.77 There were no differences in rates of nausea, vomiting, or the intensity of any nausea during the early postoperative period to 24 hours. However, the rates of vomiting after etomidate were higher (27% vs 10%).

Because it causes cortisol suppression, etomidate by infusion is contraindicated in many countries including the United States, and its use in that manner must be considered as off-label.

Midazolam

Midazolam is the only currently available benzodiazepine suitable for induction of anesthesia. It has a faster onset and lower incidence of venous complications than either diazepam or lorazepam, but a slower onset than the other IV hypnotic agents (30-60 seconds for loss of consciousness). Induction of anesthesia (as is the case with the barbiturates) coincides with loss of the eyelash reflex. Induction occurs with doses of 0.1 to 0.2 mg/kg midazolam given over 20 to 30 seconds; smaller doses are sufficient in premedicated patients, in the elderly, and in patients of ASA groups PS-III to PS-IV. However, emergence from midazolam-induced anesthesia may be prolonged.78

Midazolam has some advantages over thiopental, including improved perioperative amnesia and hemodynamic stability. However, the combination of midazolam and an opiate can result in significant cardiovascular depression. Although a single induction dose of midazolam does not suppress the adrenal steroidogenesis,79 data from Crozier et al80 and Desborough et al81 suggest that high doses of midazolam given by infusion prevent the increase in plasma cortisol concentrations in response to surgical stress. As with etomidate, the significance of this finding during TIVA is uncertain in the absence of appropriate outcome data.

Another advantage of midazolam over other hypnotic agents is the availability of a specific antagonist, flumazenil, which is devoid of any intrinsic effects on the respiratory or cardiovascular systems. However, it has a short duration of action, with peak effect occurring 1 to 3 minutes after IV injection. Because of its shorter elimination half-life (40-70 minutes) and faster clearance rate than midazolam, flumazenil may produce an initial arousal from sedation or anesthesia, followed by resedation.82 Thus, if larger doses of midazolam are used for induction and/or maintenance of anesthesia, flumazenil preferably should be given as the combination of a loading dose and continuous infusion to maintain a plasma drug concentration in excess of 20 to 40 ng/mL.

Propofol

This sterically hindered alkyl phenol was originally formulated as an emulsion containing soybean oil and egg phosphatide (Diprivan).

Propofol (as Diprivan) has a neutral pH of 7.4 and pKa of 11.0; this means that the drug is 99.7% nonionized and highly lipid soluble at pH 7.0. It is rapidly broken down to inactive metabolites (the glucuronide and the corresponding quinol glucuronide and sulphate) in the liver and possibly in other organs, such as the lungs.83,84 Other minor metabolites detected in the urine include 2-(ω-propanol)-6-isopropylphenol and 2-(ω-propanol)-6-isopropyl-1,4-quinol. It is not known whether any of these have anesthetic potencies. The hydroxylation of propofol by hepatic cytochrome P450 (CYP) involves the isoform 2B6, which shows wide individual variability.85 Another isoform that shows high binding for propofol is CYP 2C9.86

Despite its rapid and complete offset of effect, propofol has a long elimination half-life (of up to ≥45 hours), an apparent volume of distribution of 1000 to 3940 L, and a systemic clearance between 1.0 and 1.8 L/min. Because of its high extraction ratio, propofol shows flow-dependent clearance, and its own clearance is reduced secondary to its action on myocardial contractility and cardiac output. Population kinetic analyses by Schuttler and Ihmsen87 indicate weight to be a significant covariate for the elimination clearance, the 2 intercompartmental clearances of a 3-compartment model, and the apparent volumes of the central and 2 peripheral compartments, whereas in older patients (age >60 years), both elimination clearance and the central volume compartment decrease linearly with age.

Induction of anesthesia with propofol is smooth and associated with a low incidence of excitatory side effects. Doses of 1 to 2.5 mg/kg (depending on patient age, physical status, and use of premedicant drugs) induce anesthesia in approximately 30 seconds; however, lower induction doses should be used in patients with cardiovascular disease. Loss of consciousness occurs at propofol concentrations between 2.5 and 5.5 μg/mL, with the concentration for loss of response to verbal command being between 2.5 and 3.5 μg/mL. The speed of onset and the dose of propofol needed for induction are dependent on the administration rate. Stokes and Hutton88 compared induction times when propofol was infused at 50, 100, and 200 mg/min and as a 2-mg/kg bolus. Times to loss of consciousness were 124, 92, 62, and 32 seconds, respectively, with the corresponding total induction doses being 1.4, 1.96, 2.61, and 2.15 mg/kg.

The size of the induction dose depends on many physiologic factors. Kazama et al89 determined 4 factors to be independent variables influencing the size of the induction dose (namely age, lean body mass, central blood volume, and liver blood flow). When the dynamics of propofol are related to age, 3 functions correlate in a linear manner: the blood–brain equilibration rate constant and time to peak effect, the steepness of the concentration–response relationships for EEG activation and depression, and the effect-site concentration associated with 50% of peak EEG activation.90

The hemodynamic effects after induction doses of propofol are similar to those of the thiobarbiturates, although propofol additionally decreases systemic vascular resistance. In contrast to the barbiturates, propofol has little effect on heart rate. In contrast to the barbiturates, propofol has little effect on heart rate, the combination of the drug's vagotonic effect and the fall in vascular resistance predisposing to significant falls in blood pressure when used in the hypovolemic patient or in patients receiving other vagotonic drugs (eg, opiates).91 The ventilatory effects of propofol are comparable to those of other hypnotic agents. Induction doses cause significant decreases in tidal and minute volumes, coupled with episodes of apnea often greater than 30 seconds, in 25% to 30% of patients. Comparative studies indicate the duration of apnea after propofol is longer than with thiopental.92 All induction agents decrease the rib cage and abdominal components of ventilation to a similar amount. The ventilatory depressive effects of propofol are synergistic with those of midazolam and fentanyl and with the combination alfentanil-midazolam.

One advantage of propofol as an induction agent is the greater depression of pharyngeal and laryngeal reactivity than is seen with thiopental, methohexital, or etomidate. This can be of benefit during upper airway instrumentation and insertion of the laryngeal mask airway.

Pain on injection is commonly observed during induction of anesthesia with propofol, with incidences of up to 50% when administered into the small veins on the dorsum of the hand. The incidence may be reduced by use of large veins, by mixing with lignocaine (10-20 mg) and by pretreatment with drugs such as fentanyl and alfentanil. Other side effects of induction include excitatory myoclonic phenomena and occasional epileptiform fits reported during recovery. Like etomidate, propofol appears to have in vitro potential for inhibition of adrenal steroidogenesis, but this is not relevant in clinical anesthetic practice.

Ketamine (2-O-Chlorophenyl-2-Methylaminocyclohexanone HCL)

This phencyclidine derivative is formulated as a racemic mixture. It is unique among the induction agents in that it produces both dose-related unconsciousness and analgesia. After induction with ketamine, patients have little recall of surgery or anesthesia.

The 2 stereoisomers, R(–) and S(+), show differing anesthetic potencies (1:3 to 4), but have similar kinetics. The pH of ketamine is 3.5 to 5.5, and the pKa is 7.5. At physiologic pH, ketamine is highly lipid soluble, with 12% to 35% being plasma protein bound and 44% nonionized. Recovery from ketamine anesthesia occurs as a result of both distribution and degradation by demethylation and hydroxylation by hepatic cytochrome P450 2B6 and 2C9. The metabolic breakdown of ketamine is complex, but 1 metabolite (norketamine; metabolite I) is pharmacologically active with a potency of approximately 30% that of the parent drug and a longer elimination half-life. The main excretory metabolites are ketamine and metabolite I and II glucuronides. The efficacy of ketamine is enhanced in patients with renal impairment with an accompanying delayed recovery. Most of the ketamine is excreted in the urine as the glucuronides; only 2.5% is unchanged.

Ketamine causes rapid induction of anesthesia by the IV route. However, its cardiovascular effects differ from other hypnotic agents; ketamine causes increases in heart rate, blood pressure, and cardiac output. These inotropic and chronotropic effects are seen in both healthy patients and in those with heart disease and are mediated via central mechanisms. Ketamine blocks the reuptake of noradrenaline by both the uptake 1 (extraneuronal) and uptake 2 (intraneuronal) mechanisms, as well as causes the release of noradrenaline from the sympathetic ganglia. The in vivo effects are obtunded in patients receiving adrenergic antagonists (both α and β) and vasodilators, as well as by pretreatment with benzodiazepines.

Ketamine has advantages over propofol and etomidate because it is water soluble. However, although it lacks the cardiorespiratory depressive properties of other IV agents, its usefulness is limited by high incidences of disturbing emergence reactions (up to 30% of patients). Ketamine also causes increases in intracranial pressure and salivation. The psychomimetic effects of ketamine may be attenuated by benzodiazepine premedication, although these drugs appear to prolong the elimination half-life of ketamine and increase its recovery time.

Stereoisomers of Ketamine

The potency ratio for anesthesia is approximately 4:2:1 for S(+) ketamine, ketamine racemate (ie, mixture), and R(–) ketamine enantiomers. Some data suggest that administration of S(+) ketamine is accompanied by a lower median EEG power spectrum, a greater rise in blood pressure and heart rate, decreased locomotor activity, shortened recovery times, and equipotent analgesia compared with the racemate at similar doses.93-96

Opiates

As already mentioned in the section on induction of anesthesia with TIVA, these are not complete IV hypnotics and do not produce anesthesia in the absence of other supplements. However, all opiates potentiate the effects of hypnotics when used for induction of anesthesia (this is seen less with thiopental, but is most important with midazolam).

In modern anesthesia, the dosing of both hypnotics and analgesics is by titration to clinical effect, measured either by effects on the cardiovascular system or the EEG (or one of its surrogates). The changes in heart rate and blood pressure tend to be "agent specific" (for most IV agents, increasing depth of anesthesia causes a reduction in heart rate and blood pressure, although with ketamine, the heart rate may increase with increasing plasma drug concentrations). However, of all the markers of inadequate anesthesia, patient movement remains the most reliable.

Despite the availability of pharmacokinetic-based infusion regimens, it is the dynamic responses of the individual patient to any given surgery that governs the rate of drug infusion. No single plasma drug concentration results in satisfactory anesthetic and surgical conditions for all patients and all operations.

Titration of the infusion rate should reflect the anticipated intensity of the applied stimulus and likely observed patient responses. In general, drug requirements are greatest during endotracheal intubation and decrease during surgical preparation and draping. The infusion rates will need to be increased before skin incision, whereas during anesthesia, drug dosing should be titrated according to signs of patient movement, hemodynamics, and autonomic responses. In the absence of any response over a given period of time, the anesthesiologist may consider reducing the infusion rate by 15% to 20%.

Table 43-8 shows typical infusion regimens used to achieve the steady drug concentrations required to provide analgesia using IV analgesic drugs given by infusion. If the dose of drug administered is clearly too high in the presence of continuing signs of inadequate anesthesia, then the anesthesiologist should examine for a disconnection of the delivery system or delivery to a subcutaneous rather than vascular site. Other causes could include incorrect programming of pumps or mechanical errors of the delivery systems.

During TIVA, use of combinations of drugs poses questions over which one to increase or decrease and for what reasons. In general, the dosing of opiates should be aimed at achieving analgesic drug concentrations at the effect site, whereas the hypnotic infusion should be titrated to individual patient requirements and to the intensity of the surgical stimulation. At the end of surgery, the anesthesiologist should reduce the infusion rates of the hypnotic and analgesic during skin closure to allow restoration of spontaneous respiration by the end of surgery.

Thiopental and Thiamylal

Early recovery from thiopental occurs because of a decline in the blood (and brain) concentrations as a result of drug redistribution. After bolus doses and after short or low-dose infusion regimens, thiopental is eliminated by first-order kinetics, and the patient promptly awakens. However, at rates in excess of 300 μg/kg/min, thiopental concentrations increase nonlinearly because of the peripheral tissue stores becoming saturated. Maintaining anesthesia with infusions of thiopental requires rates of 150 to 300 μg/kg/min in combination with an opiate. This will achieve thiopental concentrations of 15 to 25 μg/mL.27 In the absence of an analgesic supplement, thiopental concentrations of the order of 40 to 50 μg/mL are needed to abolish the response to squeezing the trapezius muscle (which has been equated with the initial surgical incision). Besides the changes in pharmacokinetics, high doses of thiopental also lead to the formation of significant blood concentrations of its active metabolite, pentobarbital. Other inactive metabolites occur after C5 side-chain oxidation. Renal excretion of thiopental is very low (approximately 0.3%).

Possible advantages to the use of thiopental infusions include minimal cardiovascular depression and cerebral protection during ischemic episodes, with blood thiopental concentrations on the order of 70 μg/mL resulting in EEG burst suppression.97

When given by continuous infusion for TIVA, thiopental has a low systemic clearance (2.2-5.5 mL/kg/min) and an apparent volume of distribution at steady state of 1.3 to 2.4 L/kg.26,98-100 Larger volumes of distribution, coupled with altered kinetics, have been observed after prolonged or high-dose infusions.97

Methohexital

Plasma methohexital concentrations of 3 to 4 μg/mL result in hypnosis, and concentrations between 10 and 12 μg/mL cause EEG burst suppression. Based on dose–response data from Sear and Prys-Roberts,101-103 infusion rates of 50 to 65 μg/kg/min supplemented by opiates or of 100 μg/kg/min methohexital alone are required for anesthesia. Methohexital infusions depress both blood pressure and cardiac output; they also decrease baroreceptor reflex sensitivity with a resetting of the response to allow a more rapid heart rate at lower arterial pressures than when awake.104 Side effects include excitatory movements, pain on injection, and predisposition to convulsions. Epileptiform activity has been recorded by EEG, but clinical fitting is rare. Methohexital also causes pain if accidentally injected into arteries, but unlike thiopental, this does not normally lead to thrombosis.

The combination of methohexital and opiates can cause significant respiratory depression. No untoward effects of methohexital infusions on liver, renal, or adrenal function have been described.

Midazolam

Continuous infusions of midazolam have been used to provide both sedation and maintenance of anesthesia. When used as the hypnotic component to supplement alfentanil in a TIVA technique and compared with propofol, Vuyk et al observed similar hemodynamic effects but slower recovery.105 For the maintenance of anesthesia, infusions on the order of 10 mg/h (resulting in plasma drug concentrations of 200-350 ng/mL) are needed to supplement opiate infusions.7,16,78,80,106 There are few data describing the disposition of midazolam during TIVA for noncardiac surgery; typical clearance estimates range from 5.0 to 11.0 mL/kg/min and distribution 1.3 to 1.7 L/kg.107-110

Several of these studies showed that infusions of midazolam and opiates together may cause myocardial depression; this does not appear to be dose dependent, and there is an apparent ceiling effect at midazolam concentrations greater than 100 ng/mL. Clinically relevant rates of infusion of midazolam in combination with alfentanil have no significant effect on the plasma cortisol response during lower abdominal surgery.111 A major disadvantage of midazolam for TIVA is the slow recovery. Awakening occurs at drug concentrations of approximately 50 to 80 ng/mL. Nilsson et al82 showed that recovery can be improved by reversal with bolus doses of flumazenil, but re-sedation may subsequently occur as a consequence of the faster elimination of the antagonist and rebinding of the agonist.

Ketamine

Although widely used throughout the world for the maintenance of anesthesia, there are few concentration-effect data for continuous infusions of ketamine. Most studies suggest that hypnotic and analgesic thresholds are approximately 1.5 to 2.5 μg/mL and 150 to 200 ng/mL, respectively. Awakening from anesthesia occurs in the concentration range 600 to 1100 ng/mL. As sole agent, infusion rates of 60 to 80 μg/kg/min will provide clinical anesthesia.

The use of TCI ketamine for sedation and maintenance of anesthesia was described by Bowdle et al112 and Gray et al.113 For examination of the psychomimetic effects of ketamine, Bowdle et al designed a BET delivery scheme based on the kinetic parameters of Domino et al114 to achieve a stepwise series of plasma target concentrations between 0 and 200 ng/mL. Bowdle et al showed a good correlation and relationship between the targeted and observed drug concentrations. Increasing ketamine concentrations were associated with greater psychedelic effects for a variety of symptoms. Interestingly, there was no apparent threshold concentration and no concentration at which these effects plateaued.112

In the second study, Gray et al113 developed a TCI scheme for ketamine to provide the analgesic component of a TIVA technique with propofol in spontaneously breathing patients undergoing body-surface surgery, using the kinetic parameters of Wieber et al.115 The target concentration of ketamine was 300 ng/mL, with propofol delivered according to the manual BET scheme of Roberts et al.14 Clinically this combination provided good cardiovascular control throughout the surgery, with an average end-tidal carbon dioxide of 5.8 kPa. There were episodes of involuntary movements (although these did not interfere with the surgery being undertaken) and no episodes of recall. However, recovery to giving date of birth was prolonged in some patients. There were no reports of unpleasant dreams or other psychomimetic side effects.

Schuttler et al116 described a similar technique in patients undergoing lower abdominal surgery under propofol-ketamine anesthesia. They describe satisfactory anesthesia without any significant psychic disturbances or cardiovascular stimulation. Recovery was not significantly delayed.

A major development in the use of ketamine as part of TIVA has been the separation of the hypnotically active S(+) ketamine enantiomer from the racemic mixture.94 In a crossover volunteer study, White et al96 examined the pharmacologic effects of infusions of racemic, S(+), and R(–) ketamine, with measurement of cardiovascular parameters, the raw EEG, and a battery of psychometric tests. S(+) ketamine was approximately 2 times more potent in terms of anesthesia and was associated with faster recovery compared with both the racemic mixture and the R(–) isomer. This is in agreement with subsequent kinetic studies showing inhibition of metabolism of S(+) ketamine by its R(–) isomer.117 Ketamine enantiomer concentrations at time of regaining consciousness and orientation are consistent with an S-to-R potency ratio of 4:1, whereas for impairment of psychomotor function, the ratio was between 3:1 and 5:1. When administered at equipotent doses, S(+) ketamine produces longer hypnosis than the R(–) isomer, with the racemate being intermediate. Improved recovery was seen after S(+) ketamine by infusion compared with the racemate. However, cardiovascular stimulation and psychotomimetic effects are seen with both stereoisomers.

When the racemate is given by infusion, the terminal or elimination half-life of ketamine is 2.2 to 3.5 hours and clearance 14.0 to 20.0 mL/kg/min.118-120 Similar values have been described for the S+ enantiomer.121,122

When given alone as part of an anesthetic technique to surgical patients, ketamine can cause considerable side effects. Emergence reactions are commoner after infusions of the R(–) enantiomer than after the racemate and S(+) isomer. However the incidences of dreaming with all 3 treatment groups were comparable.123 There are no apparent differences between the enantiomers and racemate in their hemodynamic effects, but recovery is faster after the S(+) isomer. The EEG effects of both racemic ketamine and the S(+) isomer are similar; both cause increased fast activity (21-30 Hz) with an accompanying reduction in Δ power. The IC50 (concentration that inhibits 50%; in this instance, the plasma ketamine concentration necessary to achieve a 50% depression of the maximal EEG median frequency reduction) was 0.8 μg/mL for S(+) ketamine, compared with 1.8 and 2.0 μg/mL, respectively, for the R(–) and the racemic preparations.95 The concentration–effect relationships show the curve for S(+) ketamine to lie to the left of the racemate and to be steeper.

Propofol

Propofol is the most widely used hypnotic for the maintenance of TIVA. Infusion rates of 2 to 10 mg/kg/h are needed when administered with bolus doses or an opiate infusion, whereas drug concentrations > 8 μg/mL will be necessary if propofol is used as a sole anesthetic agent. Recovery occurs rapidly after cessation of an infusion at blood concentrations of approximately 1.0 μg/mL.91 Because of the wide variability in the therapeutic drug concentration window (related to both age and type of surgery) and intersubject drug kinetics, propofol dosing must be titrated to effect. This is easily achievable as it has a short blood–brain equilibration time (t½ke0).

The disposition of propofol has been studied extensively during TIVA or for sedation during regional anesthesia.48,87,124-126

A number of other formulations (apart from Diprivan) are either available or undergoing clinical evaluation (Table 43-9). The efficacy of these newer formulations has not been evaluated in large outcome studies, but there appears to be little difference in their kinetic and dynamic properties.

When used to maintain anesthesia, infusions of propofol cause dose-related decreases in blood pressure, cardiac output, and systemic vascular resistance.91 One important difference when compared with methohexital is that infusions of propofol do not show the normal baroreflex increase in heart rate to a decreased blood pressure. Propofol causes a resetting of the baroreceptor reflex such that slower heart rates are seen for a given arterial blood when compared with awake values.91,138 As well as these central hemodynamic effects of propofol by infusion, other cardiac effects, including severe bradycardia, sinus arrest, heart block, and asystole, have been reported, which usually occur when propofol is coadministered with vagotonic drugs.91,132 During TIVA, propofol affects ventilatory control, causing a reduction in the ventilatory response to carbon dioxide and in the acute ventilatory response to isocapnic hypoxia.139,140

Although bolus doses of propofol have no effect on renal or portal venous blood flows, dose-related changes in liver blood flow have been reported in dogs during graded infusions to concentrations greater than those needed clinically in people.141 However, in patients, clinically relevant infusion rates of propofol appear to cause no significant changes of liver blood flow or liver function tests.142

Propofol–Drug Interactions during TIVA

Because TIVA normally depends on the coadministration of more than 1 IV drug, there is always the potential for drug–drug interactions, which may result in changes to drug distribution, metabolism and elimination, or drug dynamics. In vitro, propofol inhibits drug metabolism at clinical concentrations.143-145 The magnitude of this inhibition varies from 30% to 71%, with the greatest effect on biotransformations mediated by hepatic isocytochrome P450 2B1. Other cytochromes (P450 1A1 and 2A1) are also inhibited, and Chen et al146 showed that clinical concentrations of propofol inhibit renal monooxygenase and defluorinase activities. There are 4 types of drug interaction in vivo between propofol and opioids. Besides the interaction of propofol and cytochrome P450 to inhibit drug metabolism, there is also competition between propofol and fentanyl for pulmonary binding sites.

As infusions of propofol decrease liver blood flow,141 they also decrease clearance of flow-dependent drugs. Capacity-limited drug clearance is also decreased by action of propofol on the hepatic extraction ratio.33 These findings are relevant during TIVA, because propofol may decrease the systemic clearance of other coadministered drugs through changes in effective liver blood flow or in the hepatic extraction ratio. However, simulation studies by Schnider et al21 suggest that significant changes will not be associated with propofol infusion rates used in clinical practice, as the kinetics of propofol appear linear with regard to infusion rate at those concentrations. Schnider's simulations are in general agreement with the data of Sear and colleagues141 in an open-chested dog model.

In vivo there are significant interactions between propofol and alfentanil that lead to alterations in drug clearance.147 Pavlin et al148 found that an infusion of alfentanil caused an increase in the predicted propofol concentration and led to greater than predicted alfentanil concentrations when compared with those when alfentanil was infused alone. In male volunteers, Mertens et al147 showed that addition of a propofol infusion (at a steady-state concentration ranging between 0.85 and 1.75 μg/mL) to alfentanil infused at 25 μg/kg/h resulted in a decrease in mean arterial pressure. In turn, this influenced the disposition of the alfentanil, resulting in a 15% decreased systemic clearance, a 68% reduction in the rapid distribution clearance, and a 51% reduction in slow distribution clearance.

A similar interaction is seen with the combination propofol-remifentanil, where there is a significant reduction in the initial volume of distribution of the opiate and reductions in both systemic clearance and the intercompartmental distributional clearance. Because this effect is not concentration dependent, it does affect the dosing strategy when using this opiate.19,149

Of greater importance, however, are the dynamic interactions between the hypnotics and opiates during the maintenance phase of anesthesia. One of the earliest studies examining the hypnotic–opiate interaction was by Smith et a49 (Fig. 43-2). Other clinical studies and computer simulations by Vuyk et al50,150-152 from Leiden, the Netherlands, confirm a synergism between these groups of drugs. They also indicate that regardless of an opiate's relative potency, the optimal effect-site concentration should be one that (1) prevents responses to noxious stimulation and (2) allows the rapid recovery of spontaneous ventilation at the end of anesthesia and surgery. Although the original interaction studies focused mainly on the combination propofol-alfentanil,147,148,150-152 more recent studies confirm similar interactions for propofol with remifentanil and sufentanil.39,149 A further dynamic interaction is seen between the 2 groups of drugs, with the frequent need for vasoconstrictor drugs to correct hypotension after the induction of anesthesia.

Adverse Effects of Propofol during TIVA

Although widely used as the main hypnotic component of TIVA, propofol exhibits a number of significant adverse effects, including pain on injection (especially when given into small veins and to children), hypotension and bradycardia (which are exaggerated in the presence of other vagotonic drugs, such as opioids and hypovolemia), apnea in up to 40% of patients after induction, and reports of epileptiform movements and true convulsions.

Nonhypnotic Effects of Propofol

In addition to its anesthetic effects, propofol has other properties that may be advantageous. These include:

1. Mood-altering effects: Subhypnotic doses of propofol administered by a patient-controlled analgesia (PCA) system (10 mg with a 1- to 5-minute lockout) exert sedative and anxiolytic effects in anxious patients presenting for ambulatory surgery and thus can be useful as premedication in ambulatory patients.153

2. Antiemetic effects: Although several authors have suggested an antiemetic effect of propofol (in both hypnotic and subhypnotic doses), the site of this action of the drug remains uncertain.154 It is probable that propofol does not act at dopaminergic receptors. Subhypnotic infusions of the hypnotic are also effective in the prevention of nausea and emesis after cisplatin chemotherapy.154

3. Antipruritic effects: Again, subhypnotic doses (10-20 mg IV) of propofol are equally as effective as naloxone in relieving pruritus caused by both epidural and spinally administered opiates.155,156

4. Effects on the cerebral circulation and metabolism: Although in vitro studies demonstrate a vasodilating effect of propofol, in vivo measurement shows that infusions of propofol act to decrease cerebral blood flow and intracranial pressure (ICP) and decrease the cerebral metabolic rate. Infusions of propofol have no effect on cerebrovascular autoregulation to carbon dioxide, although the slope of the curve is decreased. There is some evidence for a cerebral protective effect of propofol, and high doses of propofol have been used to afford protection during cerebral aneurysm surgery in patients requiring cardiopulmonary bypass and deep hypothermic arrest, as well as in patients undergoing nonpulsatile bypass for cardiac surgery.157,158

α2-Agonists and Other Sedative Drugs

α2-Agonist drugs are widely used as part of TIVA techniques in veterinary practice; however, few of these agents are presently licensed for clinical use for sedation or hypnosis in people. Dexmedetomidine (DMD) is presently undergoing evaluation for both its sedative and hypnotic properties. It also has analgesic properties in animals.

In humans, DMD interacts with both opiates and hypnotics to reduce drug requirements needed to maintain sedation or anesthesia. Peden et al159 examined the combination DMD and propofol. They found that an infusion of 3 ng/kg/min of DMD reduced the median effective (ED50) infusion rate of propofol for loss of consciousness from 5.79 to 3.45 mg/kg/h and the resulting plasma median effective concentration (EC50) from 2.3 to 1.69 μg/mL. However, at these infusion rates, DMD caused significant side effects (with 2 cases of sinus arrest and 1 case of severe postural hypotension persisting for 24 hours postanesthesia). Sinus arrest has also been observed in other studies. Consequently, it is appropriate to pretreat patients (especially those younger than 40 years) with an anticholinergic agent. At an infusion rate of 3 ng/kg/min, DMD blunted the hemodynamic responses to intubation and surgical incision, with the blood pressure and heart rate remaining stable throughout surgery and into the recovery period.

In a separate study, Dutta et al160 determined the EC50 of propofol for loss of motor response to an electrical stimulus to be 6.63 μg/mL; when DMD was infused to a steady plasma concentration of 0.66 ng/mL, the EC50 of propofol was reduced by 41% to 3.89 μg/mL.

A number of other clinical studies describe the use of DMD as part of TIVA techniques. Ramsey and Luterman161 used variable-rate DMD infusions to provide hypnosis for surgery in patients with upper airway pathology. When infused at rates up to 10 μg/kg/h, DMD caused no respiratory depression, none of the patients experienced severe hypotension or bradycardia, and recovery was not excessively prolonged (although the patients, who were ages 50 to 66 years, required 2 to 3 hours for complete recovery).

Clomethiazole (a derivative of the thiazole part of vitamin B1) can be used by continuous IV infusion to supplement both regional and spinal anesthesia. It permits ease of titration of the depth of sedation and anesthesia, with minimal cardiovascular effects. However, widespread use is limited because of the large fluid loads that occur when administered as a 0.8% infusion in 4% glucose; increased red cell fragility when given as infusions of greater than 5%; a high incidence of peripheral thrombophlebitis, necessitating central venous administration; and minor side effects of nasal irritation and stuffiness. After prolonged infusions, recovery may be delayed.162

Opiates

When used in TIVA, the opiates act synergistically with most hypnotic agents. Studies by Vuyk et al,6,19,50,147,150-152 Smith et al,49 and Lentschener et al163 show that even very small doses of opiates can markedly reduce the requirements of the hypnotic component. These studies also demonstrate a significant "ceiling effect"—such that above a given opiate dose or plasma concentration, little further reduction in hypnotic requirement can be achieved. During TIVA, the ability to prevent autonomic responses appears to be largely dependent on increasing the amount of the opiate drug.164 Examples of "useful" interactions between the hypnotic and analgesic components of TIVA can be found in the cited work of Vuyk et al and Mertens et al (see earlier). Table 43-8 shows the typical plasma opiate concentrations needed to obtund responses to noxious stimuli during TIVA (and the associated opiate concentrations needed for adequate spontaneous ventilation in the recovery room), together with the regimens needed to achieve other drug concentrations.

The rapidity of recovery after TIVA depends on how well the clinician is able to keep the effect-site drug concentrations near to those concentrations found in the awake or spontaneously breathing subject. If intraoperative concentrations are kept at approximately 20% above those associated with wakefulness, rapid recovery will occur.

Context-Sensitive Half-Time

In 1992, Hughes et al165 offered a different approach to the anesthesiologist's understanding of the recovery profile after infusions of intravenous anesthetic agents. Rather than relating recovery to elimination half-life or systemic clearance, they defined a new term: context-sensitive half-time (CSHT). Compared with the various half-lives of a kinetic model, the CSHT describes what is happening to the plasma drug concentration.

Termination of drug effect is dependent on 2 separate kinetic processes: drug distribution and drug elimination. An IV drug with a short elimination half-life is not necessarily a short-acting drug, as drug redistribution can affect the dynamic profile of the agent. The CSHT is dependent on the 3 rate constants—kel, k12, and k21—of a 2-compartment mamillary model in which drug elimination occurs only from the central kinetic compartment. The CSHT is a "half-time" rather than "half-life," as the time needed for plasma or effect-site drug concentration to decrease by 40% will not be twice that required to decrease by 20%! The term context refers to the duration of the drug administration, with the half-time measured from the moment of cessation of the infusion (Fig. 43-3). One limitation of the CSHT is that it describes the time to a 50% decrease in the plasma or central compartment drug concentration. This may not be the required decrement in the drug concentration needed to achieve recovery, although the studies of Prys-Roberts and Sear10 suggest that it may certainly be appropriate when predicting the behavior of hypnotic agents given by infusion.

The study of Kapila et al166 comparing the modeled CSHT for both alfentanil and remifentanil with values determined from volunteers where the effect measured was that of depression of minute ventilation validated the predictive accuracy of the CSHT. The modeled values of the CSHT after a 3-hour infusion were 2.0 and 51.9 minutes, respectively, for remifentanil and alfentanil, compared with measured values of 3.2 and 47.3 minutes, respectively.

Youngs and Shafer167 further elaborated on the concept of CSHT. They produced simulations of the time courses not just for a 50% decline in the plasma drug concentration, but also for the 20% and 80% declines after varying-length infusions of 3 opiates (alfentanil, fentanyl, and sufentanil), giving rise to the 20%, 50%, and 80% decrement times. Examination of the relationship between 6 derived components of a 3-compartment mammillary model (namely V1, V2, V3, Cl10, Cl12, and Cl13) on these decrement times show that increases in the initial volume of drug distribution (V1) lead to longer decrement times. The reverse is true after increases in the clearance constant Cl10. Similarly, increases in the peripheral volumes of distribution (V2 and V3) and the intercompartmental clearances Cl12 and Cl13 lead to shorter decrement times when the infusion is of limited duration, but longer times after prolonged infusions.

For rapid recovery from TIVA, drugs with a small initial volume of distribution (V1) and high clearance (Cl10) are the most suitable. After prolonged anesthesia, the fastest recovery times are seen after infusion of drugs with small apparent peripheral volumes of distribution and small intercompartmental clearances. Because the "biophase" concentration is always related to the plasma drug concentration, the "decrement times" are also a good descriptor of drug effect except after bolus doses or very short infusions.

Bailey described another descriptor of recovery after intravenous infusions: the "mean effect time."168 The mean effect time is based on whether a patient responds to a given parameter or end point. For a given plasma drug concentration, the "probability (p) of a given drug effect" is described by a logistic regression equation of the type:

where C = drug concentration, C50 = drug concentration at which there is a 50% probability of a given effect, and g = a parameter estimate describing the steepness of the concentration–probability relationship. From this equation, Bailey calculates the mean effect time as the integral of p against time.

Whereas the 50% decrement time of Youngs and Shafer will be, by definition, the median recovery value for a given population of patients after a given dosing regimen, it will also be the mean recovery index based on the probability of that event occurring. A large steepness coefficient g indicates a steep drug concentration–response relationship. Under such circumstances, the difference between the mean effect time and the 50% decrement times will be insignificant. However, if the concentration–response relationship is flatter, recovery times will tend to be right-skewed, and the mean effect time will be significantly greater than the 50% decrement time.

Although determination of decrement times is straightforward, the calculation of the 50% value for the intraoperative maintenance drug concentration or determination of the mean effect time requires the anesthetist to have a clear understanding of the nature of the concentration–response relationship (and hence be able to determine g). For low values of γ, the anesthetist needs to "overdose" the patient to make sure that all patients are adequately asleep or pain free. This necessitates a greater decline in the maintenance drug concentration in order to achieve concentrations where there is a 50% probability of recovery occurring. Alfentanil is an example of a drug with a high g value (9.2), whereas sufentanil, propofol, and midazolam have flatter dose–response curves (with γ values of 5.99, 3.27, and 3.05, respectively). Thus the mean effect time and the 50% decrement times for alfentanil are similar, whereas those for propofol, midazolam, and sufentanil differ, depending on the duration of drug administration, with the mean effect time being increasingly greater as the dosing time increases168(Fig. 43-4).

How Do All These Simulation Data Compare with In Vivo Observations for Different IV Agents?

Hypnotics

Propofol has a CSHT of less than 25 minutes for infusions lasting up to 3 hours duration (Fig. 43-3), and hence recovery will be prompt; in contrast, thiopental has a longer CSHT, and after infusions to achieve plasma concentrations of 10 to 20 μg/mL, recovery takes 40 to 300 minutes.100

Methohexital has a CSHT similar to that for propofol, and its recovery profile is similar to that of propofol after short-duration procedures. On the other hand, as the studies of Vuyk et al105 demonstrate, midazolam has a CSHT approximately twice that of propofol. The computed CSHT for ketamine appears to be similar to that for propofol, although its rate of elimination is slower.169 This is because ketamine equilibrates faster with the peripheral compartments. The modeled CSHT of ketamine after a BET infusion for 8 hours is approximately 50 minutes. However, its recovery profile is influenced by the high incidence of psychological reactions (seen in between 5% and 30% of patients).

Etomidate has a CSHT of approximately 50% that seen with propofol. This should lead to fast recovery after TIVA, but, again, there are associated high incidences of adverse recovery sequelae, including nausea and vomiting (15%-20% of patients), twitching and restlessness, and venous sequelae such as thrombophlebitis. The occurrence of these postoperative side effects will slow the completeness of recovery after etomidate.

Opiates

The rational use of opiates in TIVA has been well discussed by Shafer and Varvel.170 If a dosing strategy of providing analgesic concentrations is adopted, then a 20% to 30% decrease in the opiate concentration will allow adequate postoperative ventilation. Fentanyl concentrations show a rapid initial decline followed by a slower decrease, permitting spontaneous ventilation at the end of the procedure while maintaining analgesia for a considerable time thereafter. When a rapid and maximal decline in opiate concentration is desirable (as in ambulatory and day surgery), remifentanil, alfentanil, and sufentanil are more suitable (Fig. 43-5).

With infusions of less than 8 hours' duration, the decline in the sufentanil concentration is more rapid than that of alfentanil, although alfentanil has the shorter CSHT for infusions lasting 6 to 8 hours or longer. For infusions of less than 1-hour duration, fentanyl and sufentanil have CSHTs that are both shorter than alfentanil. However, for infusions of variable duration, the CSHT of remifentanil is virtually independent of its duration, with no adjustment apparently required for weight, age, or sex. Its high systemic clearance is because of its rapid metabolism by plasma esterases. The short CSHT means that most patients will achieve spontaneous ventilation within 3 to 5 minutes of stopping the infusion. However, there remains the major postoperative problem of how to provide adequate analgesia after the offset of action of the remifentanil. Although a number of strategies to overcome this have been suggested, several cases of abreaction-type responses to remifentanil have been described in the literature.164,171

These will be focused in several distinct areas. First, there is need to develop kinetic models for patient populations other than the healthy adult (ie, the young, the elderly, and those patients with comorbidities). One example is the Paedfusor for propofol administered to children.172 An alternative will be further development of existing models to allow these covariates to be taken into account (as is presently the case when they are incorporated into TCI systems that use the Schnider kinetic model for propofol delivery).21 Another major problem is obese patients, as weight alone is a poor covariate against which to determine infusion rates or target drug concentrations.

TCI technology was originally focused on propofol using the Diprifusor system. The different kinetic models for propofol questions the appropriateness of the kinetic parameters included in that system. The recent development of generic or open TCI systems allows not only use of different kinetic data sets, but also infusion of other drugs (eg, remifentanil). Whereas different equipment manufacturers produce different delivery systems, the choice of the mathematical model driving the syringe is the responsibility of the anesthesiologist. The Open TCI Initiative is an attempt to allow academics and clinicians to share data sets and develop the most appropriate parameters for given patients.

New IV drug developments will focus on those compounds with short context-sensitive half-times and that are not significantly affected by any of the major comorbidities of renal or hepatic impairment, heart failure, or diabetes mellitus. Ideally, future new IV hypnotic drugs will possess minimal cardiovascular and respiratory depression, no pain on injection, and be water soluble. However, recent studies modeling the pharmacophores for different end points of intravenous anesthesia suggest that it may not be possible to separate immobilizing activity and cardiovascular depression.173 New analgesics should not cause nausea and vomiting or prolonged or delayed respiratory depression.

Drugs fulfilling these desired kinetic and dynamic properties will either be reformulations of present compounds or new esterase-metabolized compounds similar to remifentanil.

New Propofol Formulations

Any new formulation should aim to improve the drug's profile by reducing the incidence of pain on injection; using a less toxic solvent than intralipid (which has been incriminated with the development of the so-called propofol infusion syndrome after prolonged infusion in the ICU); showing a faster onset of effect; or reducing the risk of bacterial contamination when the drug is given by continuous infusion.126 The support of bacterial growth can be blunted by the addition of benzyl alcohol, meta-bisulphite, or EDTA. The pain on injection may relate to the compound itself, as the lipid solvent is not associated with pain when used to solubilize other agents or when given alone.174

Most new formulations of propofol have focused on lipid solubilization. The Lipuro formulation (B Braun) is made up of medium-chain rather than long-chain triglycerides, has a similar PK-PD profile to Diprivan, and is associated with less pain on injection. This formulation is not presently available in the United States because of a lack of EDTA in the solvent. IDD-D propofol is another medium-chain triglyceride formulation but is prepared as a 2% solution. Studies have shown a slower onset of effect compared with Diprivan and increased pain on injection.128

Current propofol formulations are manufactured with an oil droplet size of 0.15 to 0.5 μm (fine macroemulsions)132; smaller microemulsions (<0.1 μm) are highly stable but require added surfactants. Aquafol (Daewon Pharmaceutical Co. Ltd, Seoul, South Korea) is such a formulation, containing polyethylene glycol 600 hydrostearate as a non-ionic surfactant, with 5% glycofural (Roche, Basel, Switzerland) as a co-surfactant.130,131 Its pharmacologic profile is similar to that of Diprivan; however, there are issues over the safe maximum-tolerated daily dose of both the surfactant and co-surfactant, which might limit its use in the ICU.

Three aqueous formulations of propofol solubilized in cyclodextrans have been described, with the hydroxypropyl-β cyclodextrin having the most favorable physicochemical and biologic properties.133 Studies in the pig show a similar PK/PD profile to Diprivan134; studies in man are awaited.

Other propofol formulations use co-solvent mixtures (propofol solubilized in propylene glycol: water; or the prolinate ester of propofol and its water-soluble derivative dissolved in water in equimolar concentrations). A polymeric micelle formulation has also been described,135 but again, there are no data to date in man.

The development of propofol prodrugs has been studied in laboratory animal, based on para-substituted propofol analogues. A phosphate ester formulation looks the most promising.136,137

Fospropofol

This is the only clinically approved propofol analog. It is the water-soluble phosphono-ester of propofol (phosphono-O-methyl-2,6-diisopropylphenol) (Fig. 43-6). It is broken down to propofol, inorganic phosphate, and formaldehyde by alkaline phosphatases that are widely distributed in the body. There are no reports of pain on injection, but some subjects have described transient and painful sensations in the perineal region.

In phase III studies, doses of 6.5 mg/kg of fospropofol followed by subsequent titrated doses have been used to provide procedural sedation, with fentanyl 50 μg given prior to dosing.53 Recovery was rapid and without any noticeable hangover effect. The drug has also been associated with a low incidence of postoperative nausea and vomiting. Comparison with propofol shows the new drug to have a lower incidence of adverse effects, with only minor respiratory depression. It has a slower onset of action than propofol, and the metabolic products cause no clinically significant drug-related sedative effects.

Fospropofol has some apparent advantages over propofol and is associated with a lower risk of bacterial contamination, as well as the absence of the infused lipid load that has been associated with organ toxicity during long-term infusions of Diprivan. Fospropofol was approved in the United States by the US Food and Drug Administration (FDA) in December 2008 for use in monitored anesthesia care sedation in adult patients undergoing diagnostic or therapeutic procedures. There are few comparative data relating to fospropofol and no reliable PK-PD studies. (See useful reviews by Campion and Gan175 and Garnock-Jones and Scott176 for additional information.)

Other New Intravenous Anesthetics

At the time of writing (August 2011), there are several other agents being investigated for their total intravenous anesthesia.

PF0713

PF0713 is a potent GABAA receptor agonist which, in animal species so far tested, shows a hypnotic potency comparable to that of propofol, but with a slower onset of action and duration of effect. It is a sterically hindered phenol (R,R 2′6′ di-secbutyl phenol) and is therefore similar in structure to propofol (Fig. 43-6). Indeed in its racemic form, it was one of the original compounds evaluated during the development of propofol. It is water insoluble and formulated as a 1% lipid emulsion. It is presently in phase I human volunteer clinical testing. When given to male subjects in doses of 1 to 2 mg/kg, it produces rapid loss of consciousness without injection pain or agitation. The depth and duration of anesthetic effect, as assessed by the bispectral index and the Richmond Agitation Sedation Score, was dose-related, and blood pressure and heart rate were adequately maintained.177 Although bolus dose recovery is predictable, there are no available data when given by infusion.

JM-1232 (–)

JM-1232 (–) is an isoindolin-1-one derivative and one of a series of water-soluble sedative–hypnotics presently being evaluated by the Maruishi Pharmaceutical Company (Osaka, Japan)178 (Fig. 43-6). It has a wide margin of safety, with a hypnotic ED50 of 3.1 mg/kg and LD50 of greater than 120 mg/kg (therapeutic index >35). In vitro binding data suggest that the compound has a high affinity for the benzodiazepine receptor, but not at the same binding site as midazolam. Its effect is fully reversed by flumazenil.

JM-1232 has recently undergone early clinical evaluation. When given as a 10-minute infusion to 6 male volunteers, it had a rapid onset and short duration of action.179 Doses between 0.05 and 0.8 mg/kg caused sedation, with the higher doses producing a deeper and longer reduction in the bispectral index.

Preliminary kinetic data indicate a clearance of approximately 0.78 L/min and an apparent steady-state volume of distribution of 77.3 L. Simulated context-sensitive half-times to 120-minute infusion suggest a drug with a similar profile to propofol. Further studies in a larger population of volunteers have confirmed the drug's kinetic profile and estimated an EC50 effect-site concentration for anaesthesia of 162 ng/mL (95% CI, 121-203) and a t1/2ke0 of 4.2 minutes (95% CI, 3.1-6.7).180 The latter compare with 0.9 to 5.6 minutes for midazolam and 1.2 to 3.3 minutes for propofol.

CNS 7056

CNS 7056 is an ultra–short-acting benzodiazepine hypnotic agent hydrolyzed by esterases in the blood to the metabolite 7054 (Fig. 43-6). The parent compound has a very predictable offset of effect in animals with a low risk of oversedation and is reversible by the benzodiazepine antagonist flumazenil. Metabolism of 7056 occurs rapidly by human, rat, mouse and the mini-pig liver tissue. Rapid metabolism also occurs in organs other than the liver (kidney, lung, brain), but there is no metabolism by plasma in human, mini-pig or dog. This metabolite profile is in keeping with CNS 7056 being a substrate for carboxylesterases rather than butyrylcholinesterase.181 In a kinetic and dynamic comparison with midazolam in the pig,182 both drugs rapidly induced sedation, but recovery was faster after 7056. Midazolam was eliminated more slowly (33 vs 18 min half-life) and was more widely distributed (1038 vs 440 mL/kg volume of distribution at steady state).

The first human study with CNS 7056 has been reported.183 CNS 7056 was well tolerated in doses between 0.01 and 0.35 mg/kg, although episodes of hypoxia were seen with higher doses. There was no hypo- or hypertension. Plasma drug clearance was approximately 3 times that of midazolam, and there was a linear dose-kinetic relationship. There was rapid onset of sedation (after approximately 1 minutes, with a peak at 4 minutes) and also rapid recovery (10 min vs approximately 40 minutes for equipotent doses of midazolam). A double-blind randomized controlled trial of 100 patients undergoing upper gastrointestinal endoscopy has compared midazolam and CNS 7056. Full outcome data are awaited.

THRX-918661 (Also Known as TD4756 and AZD3043)

This is another designer-built water soluble drug that is a congener of the hypnotic agent propanidid, with about twice the potency (Fig. 43-6). The structure of THRX-918661 is that of an acetic acid propyl ester broken down by tissue and plasma esterases to inactive metabolites. When given to the rat by infusions ranging from 20 minutes to 5 hours, the time to recovery of the righting reflex was approximately 3 minutes. In contrast, the recovery from propofol anesthesia ranged between 30 and 60 minutes. Faster recovery when compared with propofol has also been demonstrated in the pig.184 Kinetic studies in the pig after a 3-hour continuous infusion indicated a clearance estimate of approximately 3.4 L/kg/h and an elimination half-life of 0.4 hour, with recovery faster than after propofol at equipotent doses.185 However, the anesthetic appears to have a low potency, with doses for the maintenance of anesthesia in the pig being of the order of 1.5 mg/kg/min. Clinical studies in man are presently being conducted.

Other New Hypnotic Agents

Three further drugs (Alphaxalone formulated in cyclodextrin; MOC-etomidate; Carbo-etomidate) have all been studied in animals, but clinical trials in man are awaited.

Although Althesin in Cremophor EL continues to be used by veterinarians in some animal species, a formulation of one of the steroids (alphaxalone) in cyclodextrin is available and licensed for use in cats and dogs.

There are also a number of etomidate congeners being studied.186-189 Two potentially useful compounds are methoxy-carbonyl etomidate (MOC-etomidate) and carbo-etomidate. In a rat model, the latter appears not to cause adrenocortical suppression.186 Should this be the case in humans, the agent may be suitable for maintenance of anesthesia and/or sedation.

Other developments may include use of closed-loop control systems for the delivery of anesthesia using physiologic biomarkers such as auditory evoked potentials and the bispectral index to provide better control of the depth of sedation and anesthesia.190-194

Comparable indicators of the depth of analgesia may also be developed. It is obvious that the TIVA has become a major technique in anesthetic practice only because of the availability of drugs with potent hypnotic or analgesic effects, minimal side effects, and kinetic profiles that allow easy and rapid titration to different noxious stimuli. There are still advances to be made, including identification of the best and most accurate method of monitoring depth of anesthesia.

One of the stated advantages of TCI is the facilitation of the transition from induction to maintenance of anesthesia. However, the Cochrane Database Systematic Review by Leslie and colleagues4 does not provide any evidence showing advantages of TCI when compared with manually controlled drug delivery. Hence the striving for technological advances in drug administration in order to enable relatively accurate drug dosing should be tempered by the need to reduce the increased cost of health care delivery.

• Absalom A, Sutcliffe N, Kenny GN. Closed-loop control of anaesthesia using Bispectral Index performance assessment in patients undergoing major orthopaedic surgery under combined general and regional anaesthesia. Anesthesiology. 2002;96:67-73.
• Bailey JM. Technique for quantifying the duration of intravenous anesthetic effect. Anesthesiology. 1995;83:1095-1103.
• Garnock-Jones KP, Scott LJ. Fospropofol. Drugs. 2010;70: 469-477.
• Kazama T, Ikeda K, Morita K, et al. Comparison of the effect-site k(e0)s of propofol for blood pressure and EEG bispectral index in elderly and younger patients. Anesthesiology. 1999;90:1517-1527.
• Leslie K, Clavisi O, Hargrove J. Target-controlled infusion versus manually-controlled infusion of propofol for general anaesthesia or sedation in adults. Cochrane Database Syst Rev. 2008;3:CD006059.
• Mertens MJ, Olofson E, Engbers FHM, et al. Propofol reduces perioperative remifentanil requirements in a synergistic manner. Response surface modeling of perioperative remifentanil-propofol interactions. Anesthesiology. 2003;99:347-359.
• Mertens MJ, Vuyk J, Olofsen E, Bovill JG, Burm AG. Propofol alters the pharmacokinetics of alfentanil in healthy male volunteers. Anesthesiology. 2001;94:949-957.
• Schuttler J, Ihmsen H. Population pharmacokinetics of propofol: a multicenter study. Anesthesiology. 2000;92:727-738.
• Sear JW. Development of pharmacophoric maps for cardiovascular depression by intravenous anaesthetic agents: comparison with maps for immobilizing activity. Br J Anaesth. 2010;104:684-690.
• Struys MM, De Smet T, Depoorter B, et al. Comparison of plasma compartment versus two methods for effect compartment-controlled target-controlled infusion for propofol. Anesthesiology. 2000;92:399-406.
• Struys MM, De Smet T, Versichelen L, et al. Comparison of closed-loop controlled administration of propofol using Bispectral Index as the controlled variable versus "standard practice" controlled administration. Anesthesiology. 2001;95:6-17.
• Youngs EJ, Shafer SL. Pharmacokinetic parameters relevant to recovery from opioids. Anesthesiology. 1994;81:833-841.