Clinical Anesthesia

Chapter 19


Barbara A. Coda

Key Points

1. The term opioid designates all drugs, both natural and synthetic, including endogenous peptides, which have morphinelike properties. In its broadest sense, it refers to agonists, partial agonists, and mixed agonist–antagonists at one or more of the opioid receptors.

2. Opioid receptor classification is based on binding activity of specific ligands: morphine at mu (µ), ketocyclazocine at kappa (κ), enkephalins at delta (δ), and endorphin at epsilon (ε) receptors, and specific opioid receptors are responsible for different opioid effects. Most opioids used in clinical anesthesia today (e.g., fentanyl, morphine, and their derivatives) are highly selective for µ-opioid receptors. Naloxone, the most commonly used opioid antagonist, is not selective for opioid receptor type. Very few endogenous opioids exhibit great selectivity for a single receptor type.

3. Opioids are administered primarily for their analgesic effect, which results from complex interactions at discrete sites in the brain, spinal cord, and under certain conditions, peripheral tissues, and involves both µ1 and µ2 opioid effects. Morphine also appears to exert anti-inflammatory effects at µ3-opioid receptors. For the mixed agonist–antagonist opioids, analgesic effects are also mediated at κ receptors. Opioids act selectively on neurons that transmit and modulate nociception, leaving other sensory modalities and motor functions intact.

4. Opioids are used in combination with inhaled anesthetics to produce balanced anesthesia. Fentanyl and its derivatives reduce the minimum alveolar concentration (MAC) of volatile agents in a dose-dependent fashion. An apparent ceiling effect is seen at 70% MAC reduction, although reduction of up to 90% has been reported for sufentanil and remifentanil.

5. Fentanyl and its derivatives can be combined with a sedative-hypnotic agent to provide total intravenous anesthesia (TIVA). Alfentanil and remifentanil are particularly suited for TIVA because of their rapid onset and short duration of action.

6. Fentanyl and its derivatives can be given in very high doses for “opioid anesthesia,” but even at extremely high doses, that is, those that produce profound analgesia as well as apnea, unconsciousness is not assured.

7. Muscle hypertonus occurs with high-dose opioid administration, and severe chest wall rigidity can interfere with ventilation. It is seen most often on induction with rapid-acting opioids. Opioid-induced muscle rigidity is increased in the presence of nitrous oxide and can be prevented or treated with sedative-hypnotics or low-dose muscle relaxants.

8. All opioids depress respiratory drive in a dose-dependent manner, and ventilatory depression is seen even at doses associated with mild analgesia. Equianalgesic opioid doses produce equivalent magnitudes of respiratory depression. When given in combination with benzodiazepines, opioids can blunt hypoxic drive to a greater extent than the hypercarbic drive and produce profound respiratory depression.

9. Fentanyl or one of its derivatives is often used as a component of anesthetic induction. Small opioid doses reduce the dosage requirements of sedative-hypnotics, and blunt airway reflexes (sympathetic activity in response to laryngoscopy).

10.     In low doses, opioids have minimal cardiovascular effects, but bradycardia and hypotension are seen with higher doses. A prominent feature of fentanyl and its derivatives is their remarkable hemodynamic stability. Morphine and meperidine cause histamine release, and high doses of these opioids can produce hypotension.

11.     All opioids can produce nausea and vomiting through complex interactions at nausea and vomiting centers in the medulla. In general, equianalgesic doses of opioids produce similar magnitude of nausea. Opioid-induced nausea can be exacerbated by vestibular input, and is particularly problematic in ambulatory patients.

12.     Opioids produce smooth muscle spasm throughout the gastrointestinal tract. They decrease gastric secretions and delay gastric emptying. Opioids increase the tone of the common bile duct and sphincter of Oddi, although meperidine and the mixed agonist–antagonists cause less biliary spasm than morphine and fentanyl.



Opioids have been used in the treatment of pain for thousands of years. The drug opium, which contains more than 20 alkaloids, is obtained from the exudate of Papaver somniferumseed pods, and the word opium is derived from opos, the Greek word for juice. The first undisputed reference to poppy juice is found in the third century (BCE) writings of Theophrastus.1 The German pharmacist Sertuener isolated what he called the “soporific principle” in opium in 1806, and in 1817 named it morphine, after the Greek god of dreams, Morpheus.2 Isolation of other opium alkaloids followed, and by the mid-1800s, the medical use of pure alkaloids rather than crude opium preparations began to spread.1 Morphine was used widely to treat wounded soldiers during the American Civil War, and in 1869 its use as a premedication was described by Claude Bernard. However, in the absence of muscle relaxants and controlled ventilation, opioids were associated with a significant risk of severe respiratory depression and death. Thus their use in anesthesia was limited at that time.

With the advent of cardiac surgery in the late 1950s came the development of “opioid anesthesia.” A decade later, Lowenstein3 reported the use of progressively higher doses of morphine (0.5 to 3 mg/kg), but found limitations including incomplete suppression of the stress response, hypotension, awareness during anesthesia, and increased fluid and blood requirements.

Fentanyl, a 4-anilinopiperidine derivative of phenoperidine, was synthesized in 1960. The completely synthetic opioids were more potent and had a better safety margin (ratio of median lethal dose to lowest effective dose for surgery) than meperidine. Advances in surgical techniques created the need for potent opioids with a rapid onset, a brief, predictable duration, and a maximal safety margin for use in clinical anesthesia, and led to development of sufentanil, alfentanil, and other fentanyl derivatives between 1974 and 1976. The newest potent opioid, remifentanil, has an ultrashort duration of action because of rapid metabolism by ester hydrolysis and offers an advantage in specific clinical settings.

The search for opioid analgesics without potential for dependence was stimulated by concerns about opioid addiction and led to the identification of multiple opioid receptor types. In the mid-1960s, nalorphine, a morphine antagonist, was also found to have analgesic properties. Two other compounds, pentazocine and cyclazocine, antagonized some of morphine's effects. Pentazocine also produced analgesia, and both produced some psychotropic effects that morphine did not. These and other observations led Martin4 to propose the theory of receptor dualism. Intrinsic to this theory were two key concepts: (1) the existence of multiple opioid receptors (originally only two were proposed) and (2) the idea of pharmacologic redundancy (i.e., more than one receptor could mediate a physiologic function, such as analgesia). Thus, a drug could be a strong agonist, a partial agonist, or a competitive antagonist at one or more of the different receptor types. Subsequent research has revealed three distinct families of opioid peptides and multiple categories of opioid receptors.


The term opiate was originally used to refer to drugs derived from opium, including morphine, its semisynthetic derivatives, and codeine. The more general term opioid was introduced to designate all drugs, both natural and synthetic, with morphinelike properties, including endogenous peptides. The nonspecific term narcotic has been used to refer to morphine and morphinelike analgesics. However, because of its use in a legal context, referring to any drug (including nonopioids such as cocaine) that can produce dependence, the term narcotic is not useful in a pharmacologic or clinical context.

Figure 19-1. Log dose-effect curves for two agonists (A and B) with equal efficacy but different potency, and a partial agonist (C). Note that the potencies of A and C are similar, but the efficacy is less and the slope of the dose-response curve is shallower for the partial agonist. Note also that at lower doses, the partial agonist C is more potent than the full agonist B.

In its broadest sense, the term opioid can refer to agonists, partial agonists, mixed agonist–antagonists, and competitive antagonists. Differentiation of these terms requires understanding of receptor-ligand interactions. Receptor theory states that drugs have two independent characteristics at receptor sites: affinity, the ability to bind a receptor to produce a stable complex, and intrinsic activity or efficacy, which is described by the dose-effect curve resulting from the drug–receptor combination. Efficacy can range from zero (i.e., no effect) to the maximum possible effect, depicted graphically as the plateau of the dose-effect curve (Fig. 19-1). Given a high enough dose, an agonist will produce the maximum possible effect of binding with the receptor, whereas an antagonist produces no direct effect when it binds the receptor. A partial agonist has a dose-effect ceiling that is lower than the maximum possible effect produced by a full agonist, as well as a dose-effect curve that is less steep than that of a full agonist. A mixed agonist–antagonist acts as an agonist (or partial agonist) at one receptor and an antagonist at another. It is important to differentiate the term potency from efficacy. Whereas efficacy defines the range in magnitude of an effect produced by a drug-receptor combination relative to the maximum possible effect, potency refers to the relative dose required to achieve an effect, and is related to receptor affinity. Thus, at the lower end of the effect range, a partial agonist may be more potent than a full agonist (Fig. 19-1). However, even at very large doses the efficacy, or maximum effect achieved by the partial agonist, will be less than the maximum possible effect of a full agonist.

Endogenous Opioids and Opioid Receptors

All of the endogenous opioids are derived from three prohormones: proenkephalin, prodynorphin, and pro-opiomelanocortin (POMC). Each of these precursors is encoded by a separate gene. The three families of peptides differ in their


distribution, receptor selectivity, and neurochemical role,5 but share some features. For example, all begin with the pentapeptide sequences of [Leu]- or [Met]-enkephalin. Proenkephalin includes the pentapeptide sequences for [Met]- and [Leu]-enkephalin, and cells that synthesize proenkephalin are widely distributed throughout the brain, spinal cord, and peripheral sites, particularly the adrenal medulla.6 Pro-opiomelanocortin is the common precursor of β-endorphin, adrenocorticotropic hormone (ACTH), and melanocyte-stimulating hormone. The term endorphin is reserved for peptides of the POMC family. The major site of POMC synthesis is the pituitary, but it is also found in the pancreas and placenta. The dynorphin peptides all begin with the [Leu]-enkephalin sequence and are widely distributed throughout the brain, spinal cord, and peripheral sites.

Endogenous opioids bind to a number of opioid receptors to produce their effects. The initial classification by Martin4 of opioid receptors into the three types was based on binding activity of the exogenous ligands morphine, ketocyclazocine, and SKF10,047 at mu (µ), kappa (κ), and sigma (σ) receptors, respectively. Other opioid receptors identified since that time are delta (δ) receptors, bound by enkephalins, and epsilon (ε) receptors, bound by endorphin.5,6 There is also evidence supporting the existence of two µ, two δ, and three κ receptor subtypes.7 Increasing evidence supports a third µ receptor subtype, present on human vascular tissues and leukocytes.8 While it appears that specific opioid receptors are responsible for different opioid effects and that synthetic opioids may be highly selective for a receptor type or subtype, it is important to note that very few endogenous opioids exhibit great selectivity for a single receptor type.9

Remember also, that the theory of receptor dualism includes the concept of pharmacologic redundancy of receptor function. Thus, observed opioid effects typically involve complex interactions among the different receptor systems at supraspinal, spinal, and peripheral sites. The expression of endogenous opioids and opioid receptors is not a static phenomenon. For example, acute inflammation has been shown to up-regulate the expression of both β-endorphin and met-enkephalin as well as peripheral µ- and δ-opioid receptors.10,11Conversely, chronic inflammation is associated with down-regulation of µ-opioid receptors.12 µ-opioid receptor expression was also decreased in the dorsal root ganglia in a nerve injury model of neuropathic pain.13 Table 19-1 summarizes our current understanding of which opioid receptors are responsible for mediating opioid analgesic and side effects. One caveat in interpreting this summary is that species differences in opioid receptor systems exist, so the results of animal studies, from which most of this information is derived, may not always be directly applicable to humans. Most opioids used in clinical anesthesia today (e.g., fentanyl, morphine, and their derivatives) are highly selective for µ-opioid receptors. Naloxone, the most commonly used opioid antagonist, is not selective for opioid receptor type. In fact, current identification of an opioid-receptor–mediated drug effect requires demonstration of naloxone reversibility. Development of selective


opioid receptor subtype antagonists, such as naltrindole (a δ-opioid receptor antagonist) and nor-binaltorphimine (a κ-opioid receptor antagonist) are currently improving our understanding of which receptor subtypes mediate specific opioid effects.

Table 19-1 Tentative Classification of Opioid Receptor Subtypes and Their Actions









↓ Gastric secretion



↓ GI transit—supraspinal and peripheral


Skeletal muscle rigidity
?Urinary retention (and/or δ)




Biliary spasm (probably >1 receptor type)



Prolactin release

Acetylcholine turnover


Spinal and supraspinal (synergism with spinal)

Respiratory depression

↓ GI transit—spinal and supraspinal

Most cardiovascular effects





↓ ADH release







(Pharmacology unknown)






?Respiratory depression

↓ GI transit—spinal
Antidiarrheal—spinal and supraspinal

?Growth hormone release

?Urinary retention (and/or µ)



Dopamine turnover




Unknown (receptor type not identified)


Pupillary constriction
Nausea and vomiting

GI, gastrointestinal; ADH, antidiuretic hormone.
Adapted from Pasternak GW: Pharmacological mechanisms of opioid analgesics. Clin Neuropharmacol 1993; 16: 1.

At the cellular level, endogenous and exogenous opioids produce their effects by altering patterns of interneuronal communication. Receptor binding initiates a series of physiologic functions resulting in cellular hyperpolarization and inhibition of neurotransmitter release, effects that are mediated by second messengers. All opioid receptors appear to be coupled to G proteins,5 which regulate the activity of adenylate cyclase among other functions. G protein interactions, in turn, affect ion channels; different ion conductances may be involved at different opioid receptor types.9

Structure–Activity Relationships

The wide array of different molecules that produce morphinelike analgesia and side effects, including endogenous opioids, all share some common structural characteristics. Horn and Rodgers7 suggested that the tyrosine moiety at the amino terminal of the enkephalins formed the basis of a significant conformational relationship between the enkephalins and opiates. The structure of the phenanthrene class of opium alkaloids is complex and consists of five or six fused rings. Morphine, one of three phenanthrenes, has a rigid five-ring structure that conforms to a “T” shape (Fig. 19-2).14 The other phenanthrenes are codeine, a derivative of morphine, and thebaine, a precursor of oxycodone and naloxone. Progressively reducing the number of fused rings from the phenanthrenes yields the morphinans, with four rings; the benzomorphans, with three rings; the phenylpiperidines, with two rings; and finally, the tyramine moiety of the endogenous opioid peptides, with a single hydroxylated ring. All of these distinct classes of drugs possess morphinelike activity. The opiate receptor model of Thorpe14 is based on these structural similarities with two aromatic binding sites and one anionic site responsible for binding the positively charged nitrogen. In this model, differences in binding at the aromatic or anionic sites could account for receptor specificity or for agonist versus antagonist activity. Structural modifications alter such important properties as opioid receptor affinity, agonist versus antagonist activity, resistance to metabolic breakdown, lipid solubility, and pharmacokinetics.1

Figure 19-2. “T”-shape conformation of opioid molecules. A. Morphine, one of the phenanthrene alkaloids, has a rigid five-ring structure, with a phenylpiperidine ring forming a crossbar and a hydroxylated aromatic ring in the vertical axis. B. Reducing the number of fused rings to four yields the morphinan class of opium alkaloids. C. Benzomorphans have three fused rings. D. Phenylpiperidines and the 4-anilinopiperidines such as fentanyl have a flexible two-ring structure. E. Finally, the tyramine moiety, which is the amino terminal peptide of both [Leu]- and [Met]-enkephalin, is shown, with a single aromatic ring. Another key feature is the positively charged basic nitrogen equidistant (4.55 Å) from the aromatic ring. (Adapted with permission from Thorpe DH: Opiate structures and activity: A guide to understanding the receptor. Anesth Analg 1984; 63: 143.)

Pharmacokinetics and Pharmacodynamics

Opioid effects are initiated by the combination of an opioid with one or more receptors at specific tissue sites. The relationship between opioid dose and effects depends on both pharmacokinetic and pharmacodynamic variables. Pharmacokinetics determines the relationship between drug dose and its concentration at the effect site(s). Pharmacodynamicvariables relate the concentration of a drug at its site of action, in this case opioid receptors in the brain and other tissues, and the intensity of its effects. Pharmacokinetics generally refers to the study of blood or plasma drug concentration versus time because blood is easy to sample, bears a definable relationship to tissue concentration, and is the medium by which drugs are distributed throughout the body. Changes in drug concentration over time in the blood, at the effect site and at other sites, are determined by physicochemical properties of the drug as well as the processes of absorption, redistribution, biotransformation, and elimination.

In clinical anesthesia practice, opioids are typically administered intravenously. After an intravenous (IV) bolus dose or brief infusion, peak plasma opioid concentrations occur within minutes. Plasma drug concentrations then fall rapidly as the drug is distributed to extravascular sites, including sites of action, noneliminating tissues, and eliminating organs. Compartmental models describe the time course of change in plasma concentration; typically, opioids used in anesthesia are characterized by two- or three-compartment models (seeChapter 7). The early rapid decline in plasma concentration after the peak is called the distribution phase, and the subsequent slower decline is the elimination phase. From a mathematical curve fitted to measured plasma concentration versus time data, distribution and elimination half-lives, systemic


clearance, compartment volumes, and intercompartmental transfer rate constants can be calculated. Table 19-2 summarizes the estimates of key pharmacokinetic parameters and physicochemical characteristics for the most commonly used opioids in clinical anesthesia.

Table 19-2 Physicochemical Characteristics and Pharmacokinetics of Commonly Used Opioid Agonists in Adults















% Nonionized (pH 7.4)














Protein binding (%)







Clearance (mL/min)







Vdss (L)







Rapid distribution half-life (T1/2π, min)





Slow redistribution half-life (T1/2π, min)







Elimination half-life (T1/2β, h)







λow, octanol: water partition coefficient; Vdss, steady-state volume of distribution.
aUnpublished information from Glaxo. J. G. Bovill, personal communication, 1995.
bGlass PSA, et al: Anesth Analg 1999; 89: S7.
Adapted from Bovill JG: Pharmacokinetics and pharmacodynamics of opioid agonists. Anaesth Pharmacol Rev 1993; 1: 122.

It is important to note that there is tremendous variability in the values published for opioid pharmacokinetic parameters. This is partly because of real population differences (e.g., age, diseases) and partly because of differences in study design (e.g., sampling site, duration, concomitant events such as surgery, or other drugs that may affect differential flow to sites of metabolism or elimination). In addition, the distributional and elimination half-lives are of limited use in predicting the onset and duration of opioid action in clinical anesthesia. Contributions of distribution processes between physiologic compartments vary with time. In an effort to relate pharmacokinetics to the time of onset and duration of action, concepts such as effect compartment in pharmacodynamic modeling15 and context-sensitive half-times16 have been developed. The application of these concepts is considered later in this chapter.

Physicochemical properties of opioids influence both pharmacokinetics and pharmacodynamics. To reach its effector sites in the central nervous system (CNS), an opioid must cross biologic membranes from the blood to receptors on neuronal cell membranes. The ability of opioids to cross this blood–brain barrier depends on such properties as molecular size, ionization, lipid solubility, and protein binding (Table 19-2). Of these characteristics, lipid solubility and ionization assume major importance in determining the rate of penetration to the CNS. In the laboratory, lipid solubility is measured as an octanol: water or octanol: buffer partition coefficient. Drug ionization is also an important determinant of lipid solubility; nonionized drugs are 1,000 to 10,000 times more lipid-soluble than the ionized form.17 The degree of ionization depends on the pKa of the opioid and the pH of the environment. An opioid with a pKa much lower than 7.4 will have a much greater nonionized fraction in plasma than one with a pKa close to or greater than physiologic pH. While greater lipid solubility correlates with membrane permeability, the relationship is not simply a linear one. Hansch and Dunn18 have shown that there is an optimal hydrophobicity for blood–brain barrier penetration, and Bernards and Hill19 have demonstrated a similar biphasic relationship between the octanol: buffer distribution coefficient and spinal meningeal permeability. Plasma protein binding also affects opioid redistribution because only the unbound fraction is free to diffuse across cell membranes. The major plasma proteins to which opioids bind are albumin and α1-acid glycoprotein. Alterations in α1-acid glycoprotein concentration occur in a variety of conditions and disease states and result in acute or chronic changes in opioid requirements.

Two main mechanisms are responsible for drug elimination: biotransformation and excretion. Opioids are biotransformed in the liver by two types of metabolic processes. Phase I reactions include oxidative and reductive reactions, such as those catalyzed by cytochrome P450 system, and hydrolytic reactions. Phase II reactions involve conjugation of a drug or its metabolite to an endogenous substrate, such as D-glucuronic acid.17 Remifentanil is metabolized via ester hydrolysis, which is unique for an opioid. With the exceptions of the N-dealkylated metabolite of meperidine and the 6- and possibly 3-glucuronides of morphine, opioid metabolites are generally inactive. Opioid metabolites and, to a lesser extent, their parent compounds are excreted primarily by the kidneys. The biliary system and gut are other routes of opioid excretion.


Morphine produces its major effects in the CNS and the gastrointestinal system, but other systems are also affected. CNS effects include analgesia, sedation, changes in affect, respiratory


depression, nausea and vomiting, pruritus, and changes in pupil size. Morphine also affects gastric secretions and gut motility, and has endocrine, urinary, and autonomic effects. It mimics the effects of endogenous opioids by acting as an agonist at µ1- and µ2-opioid receptors throughout the body and is considered the standard agonist to which other µ-agonists are compared.


Morphine analgesia results from complex interactions at a number of discrete sites in the brain, spinal cord, and under certain conditions, peripheral tissues, and involves both µ1 and µ2 opioid effects. Morphine and related opioids act selectively on neurons that transmit and modulate nociception, leaving other sensory modalities and motor functions intact. At the spinal cord level, morphine acts presynaptically on primary afferent nociceptors to decrease the release of substance P and also hyperpolarizes postsynaptic neurons in the substantia gelatinosa of the dorsal spinal cord to decrease afferent transmission of nociceptive impulses.20 Spinal morphine analgesia is mediated by µ2-opioid receptors. Supraspinal opioid analgesia originates in the periaqueductal gray matter, the locus ceruleus, and nuclei within the medulla, notably the nucleus raphe magnus, and primarily involves µ1-opioid receptors. Microinjections of morphine into any of these regions activate the respective descending modulatory systems to produce profound analgesia.6,20 Endogenous pain transmission and modulation pathways are discussed in Chapter 58. Morphine can act at a number of these discrete regions in the CNS to produce synergistic analgesic effects. For example, coadministration at the level of the brain and spinal cord increases morphine's analgesic potency nearly tenfold,21 an effect mediated by µ2-opioid receptors.6 There are also synergistic interactions between supraspinal sites of opioid action (e.g., between the periaqueductal gray matter and the nucleus raphe magnus).6 When acute inflammation is present, morphine may also produce analgesia by activating peripheral opioid receptors.11,22 In chronic pain conditions such as neuropathic pain or chronic arthritis, spinal and peripheral receptors may be down-regulated, a state that can decrease morphine analgesia.12,13

Although rapidly changing plasma morphine concentrations, such as those that follow bolus dosing, do not correlate well with analgesic effects, constant or very slowly changing (i.e., steady-state) plasma concentrations do correlate with effect intensities. The minimum effective analgesic concentration (MEAC) of morphine for postoperative pain relief is 10 to 15 ng/mL.23 For more severe pain, plasma morphine concentrations of 30 to 50 ng/mL are needed to achieve adequate analgesia.24

Effect on Minimum Alveolar Concentration of Volatile Anesthetics

µ-agonists are used extensively in conjunction with inhaled anesthetics to provide “balanced anesthesia.” In animals, morphine decreases the minimum alveolar concentration (MAC) of volatile anesthetics in a dose-dependent manner,25,26 but there appears to be a ceiling effect to the anesthetic-sparing ability of morphine, with a plateau at 65% MAC.25 Morphine 1 mg/kg administered with 60% nitrous oxide (N2O) blocks the adrenergic response to skin incision in 50% of patients, a characteristic called MAC-BAR.27 Neuraxial morphine may also reduce MAC. Epidural morphine 4 mg given 90 minutes prior to incision reduces halothane MAC by nearly 30%.28 The effect of intrathecal morphine on MAC is unclear. In one study, a relatively large dose of intrathecal morphine (750 µg) reduced halothane MAC approximately 40%,29 but an equally large dose (15 µg/kg) failed to reduce halothane MAC in another.30

Other Central Nervous System Effects

Morphine can produce sedation, as well as cognitive and fine motor impairment, even at plasma concentrations commonly achieved during management of moderate to severe pain.31Other subjective side effects include euphoria, dysphoria, and sleep disturbances. High doses of morphine and similar opioids produce a slowing of electroencephalogram (EEG) activity associated with a marked shift toward increased voltage and decreased frequency.1,32 In routine analgesic doses, morphine can produce sleep disturbances, including reduction in rapid eye movement and slow-wave sleep,1 as well as vivid dreams.

Morphine produces dose-dependent pupillary constriction (miosis).33 In the absence of other drugs, miosis appears to correlate with opioid-induced ventilatory depression. However, hypoxemia from severe opioid-induced respiratory depression will cause pupillary dilation.

Systemic and neuraxial administration of morphine can produce pruritus, although this symptom is more common with spinal administration.1 Pruritus appears to be a µ receptor-mediated effect produced at the level of the medullary dorsal horn,34 although there may also be a direct antipruritic effect mediated by κ receptors.35 Antihistamines are often used to treat this side effect, but pruritus induced by morphine microinjection into the medullary dorsal horn is not histamine-mediated.34 Thus, their effectiveness is probably related to nonspecific sedative effects.

Morphine can also affect the release of several pituitary hormones, both directly and indirectly. Inhibition of corticotropin-releasing factor and gonadotropin-releasing hormone decreases circulating concentrations of ACTH, β-endorphin, follicle-stimulating hormone, and luteinizing hormone. Prolactin and growth hormone concentrations may be increased by opioids, and antidiuretic hormone release is inhibited by opioids.1

Respiratory Depression

Morphine and other µ agonists produce dose-dependent ventilatory depression primarily by decreasing the responsivity of the medullary respiratory center to CO2.33 Standard therapeutic doses of morphine produce a shift to the right and a decrease in slope of the ventilatory response to CO2 curve, as well as abnormal breathing patterns.36,37 The respiratory depressant effects of morphine are similar for young and elderly patients,36,37 but normal sleep markedly potentiates morphine-induced ventilatory depression.38 Frequent periods of oxygen desaturation associated with obstructive apnea, paradoxic breathing, and slow respiratory rate have been reported in patients receiving morphine infusions for postoperative analgesia, but occurred only when the patients were asleep.39 Such reports emphasize the need to consider both the expected severity of postoperative pain as well as diurnal variations in pain and opioid sensitivity when including long-acting opioids such as morphine in an anesthetic. Sleep apnea, often seen in association with obesity, increases the risk of morphine-induced respiratory depression. With increasing morphine doses, periodic breathing resembling Cheyne-Stokes breathing, decreased hypoxic ventilatory drive, and apnea can occur.40 However, even with severe ventilatory depression, patients are usually arousable and will breathe on command.

Cough Reflex

Morphine and related opioids depress the cough reflex, at least in part by a direct effect on the medullary cough center. Doses required to attenuate the cough reflex are smaller than the usual analgesic dosage, and receptors mediating this effect appear to be less stereospecific and less sensitive to naloxone


than those responsible for analgesia.1 Dextroisomers of opioids, which do not produce analgesia, are also effective cough suppressants.1

Muscle Rigidity

Large doses of IV morphine (2 mg/kg infused at 10 mg/min) can produce abdominal muscle rigidity and decrease thoracic compliance; this effect plateaus 10 minutes after morphine administration is complete.41 Subjects receiving smaller doses of IV morphine (10 to 15 mg) also report feelings of muscle tension, most frequently in the neck or legs, but occasionally in the chest wall (unpublished observations). Muscle rigidity is drastically increased by the addition of 70% N2O.41 Opioid-induced muscle rigidity appears to be mediated by µ receptors at supraspinal sites.42 Myoclonus, sometimes resembling seizures, but without EEG evidence of seizure activity, has also been observed with high-dose opioids.43 In clinical practice, opioid-induced muscle rigidity and myoclonus are most often observed on induction of anesthesia, but have been observed postoperatively44 and can be severe enough to interfere with manual or mechanical ventilation. These effects are reduced or eliminated by naloxone,44 drugs that facilitate γ-aminobutyric acid agonist activity (such as thiopental41 and diazepam), and muscle relaxants.44

Nausea and Vomiting

Nausea and vomiting are among the most distressing side effects of morphine and its derivatives. Increased postoperative vomiting is seen with morphine premedication as well as with the use of intraoperative opioids.45 The incidence of opioid-induced nausea appears to be similar irrespective of the route of administration, including oral, IV, intramuscular, subcutaneous, transmucosal, transdermal, intrathecal, and epidural.45 Furthermore, laboratory and clinical studies comparing the incidence or severity of nausea and vomiting have found no differences among opioids in equianalgesic doses, including morphine, hydromorphone, meperidine, fentanyl, sufentanil, alfentanil, and remifentanil.45,46,47,48 The physiology and neuropharmacology of opioid-induced nausea and vomiting are complex (Fig. 19-3). The vomiting center receives input from the chemotactic trigger zone (CTZ) in the area postrema of the medulla, the pharynx, gastrointestinal tract, mediastinum, and visual center.45,49 The CTZ is rich in opioid, dopamine, serotonin, histamine, and (muscarinic) acetylcholine receptors, and also receives input from the vestibular portion of the eighth cranial nerve. Morphine and related opioids induce nausea by direct stimulation of the CTZ and can also produce increased vestibular sensitivity.1 Therefore, vestibular stimulation such as ambulation markedly increases the nauseant and emetic effects of morphine. This can be especially problematic in outpatient surgery, when early ambulation is a clinical priority. High doses of morphine and other opioids also have naloxone-reversible antiemetic effects at the level of the vomiting center.50 In volunteer studies, morphine-induced nausea and vomiting increase after a morphine infusion is stopped,51 which suggests that antiemetic effects are more short-lived than emetic effects. Another possible explanation for this observation is that the active metabolite morphine-6-glucuronide accumulates and worsens nausea. Prophylaxis and treatment of opioid-induced nausea and vomiting includes the use of drugs that act as antagonists at the various receptor sites in the CTZ as well agents such as propofol and benzodiazepines, whose antiemetic mechanisms are unknown.45

Figure 19-3. Pharmacology of nausea and vomiting. The chemotactic trigger zone (CTZ), located in the area postrema of the brainstem, contains dopamine, serotonin, histamine, and muscarinic acetylcholine as well as opioid receptors. The vomiting center receives input from the CTZ as well as peripheral sites via the vagus nerve. As illustrated, the role of opioids is complex, and they appear to have both emetic and antiemetic effects. CN, cranial nerve; GI, gastrointestinal; n., nerve.

Gastrointestinal Motility and Secretion

Morphine and other opioids affect gastrointestinal motility and propulsion, as well as gastric and pancreatic secretions via stimulation of opioid receptors in the brain, spinal cord, enteric muscle, and smooth muscle,40,52 and are mediated by µ-, κ-, and δ-opioid receptors at different anatomic sites.52 In rodents, µ agonists inhibit gastric secretion, decrease gastrointestinal motility and propulsion, and suppress diarrhea when administered by intracerebroventricular, intrathecal, and peripheral injection.52 Animal and human studies demonstrate that methylnaltrexone and alvimopan, opioid antagonists that do not cross the blood–brain barrier, attenuate morphine-induced gastrointestinal dysfunction.53,54 Such studies also suggest that effects such as delayed gastric emptying, ileus, and constipation are mediated primarily by a peripheral opioid mechanism. Morphine decreases lower esophageal sphincter


tone and produces symptoms of gastroesophageal reflux,40 and diamorphine significantly slows gastric emptying. Thus, preoperative opioid administration should be considered when evaluating the risk of regurgitation and aspiration of gastric contents in patients who will be anesthetized or sedated. Like other opioid effects, gastrointestinal effects are probably dose-related. Tone in both the small and large bowel is increased, but propulsive activity is decreased, leading to constipation. Epidurally administered morphine can also delay gastric emptying.55

Biliary Tract

Morphine and other opioids increase the tone of the common bile duct and sphincter of Oddi. Symptoms accompanying increases in biliary pressure can vary from epigastric distress to typical biliary colic, and may even mimic angina. When produced, biliary spasm can elevate plasma amylase and lipase for up to 24 hours.1 Morphine and other µ agonists such as fentanyl are used in provocative tests to evaluate sphincter of Oddi dysfunction and biliary-type pain. In volunteers, morphine caused a greater delay in gallbladder emptying56 and an increase in contractions of the sphincter of Oddi57 than meperidine. Nitroglycerine, atropine, and naloxone can reverse opioid-induced increases in biliary pressure.1 It has been suggested that morphine causes biliary tract contraction via histamine release, and antagonism of morphine's biliary effects by diphenhydramine supports this hypothesis.58

Genitourinary Effects

Urinary retention, seen after both systemic and spinal morphine administration, is caused by complex effects on central and peripheral neurogenic mechanisms. It results in dyssynergia between the bladder detrusor muscle and the urethral sphincter because of a failure of sphincter relaxation.1,59 Estimates of the incidence of this bothersome side effect vary widely and are confounded by the effects of anesthesia and surgery on urinary retention, but it is probably more common after spinal administration. Spinal morphine appears to cause naloxone-reversible urinary retention via µ- and/or δ-, but not κ-opioid receptors.59 In an animal study, cholinomimetic agents and α-adrenergic agonists aggravated morphine-induced high intravesical pressures, and therefore may be harmful agents to use for treatment of morphine-induced urinary retention.60

Histamine Release

Opioids stimulate the release of histamine from circulating basophils and from tissue mast cells in skin and lung.61,62 Morphine-mediated histamine release is dose-dependent; intradermal injection of morphine in a concentration of 1 mg/mL induces an urticarial wheal and flare.62 Morphine-induced histamine release is not prevented by pretreatment with naloxone,62 suggesting that histamine release is not mediated by opioid receptors. Morphine-induced histamine release has clinical relevance. The decrease in peripheral vascular resistance seen with high-dose morphine (1 mg/kg) correlates well with elevated plasma histamine concentration.63 Furthermore, differences in the release of histamine could account for most of the hemodynamic differences between morphine and fentanyl (Fig. 19-4).63

Cardiovascular Effects

In doses typically used for pain management or as part of balanced anesthesia, morphine has little effect on blood pressure or heart rate and rhythm in the supine, normovolemic patient. However, higher doses of morphine can produce arteriolar and venous dilation, decreased peripheral resistance, and inhibition of baroreceptor reflexes,1 which can lead to postural hypotension. In addition to histamine release, morphine-mediated central sympatholytic activity and direct action on vascular smooth muscle may also contribute to peripheral vasodilation.64 Thus, morphine's effect on vascular resistance is greater under conditions of high sympathetic tone.57 The clinical implications of this finding are important. Patients who are critically ill (e.g., patients with severe trauma or cardiac disease) can be expected to have high sympathetic tone, and thus may experience hypotension in response to doses of morphine that would not normally produce hemodynamic instability. At clinically relevant doses, morphine does not suppress myocardial contractility.1 However, opioids do produce dose-dependent bradycardia, probably by both sympatholytic and parasympathomimetic mechanisms.65 In clinical anesthesia practice, opioids are often used to prevent tachycardia and reduce myocardial oxygen demand. Patients undergoing cardiovascular surgery who received 1 to 2 mg/kg of morphine experienced minimal changes in heart rate, mean arterial pressure, cardiac index, and systemic vascular resistance. However, outcome was no different from that achieved with carefully administered inhalation-based anesthesia.65 Morphine's


specific ability to reduce systemic inflammation by its action at the µ3-opioid receptor may benefit patients undergoing cardiopulmonary bypass.66 Murphy et al.67 demonstrated that morphine suppresses several components of the inflammatory response to cardiopulmonary bypass. Clinical benefits of morphine 40 mg, given prior to cardioplegia, include better recovery of global ventricular function and prevention of postoperative hypothermia.68

Figure 19-4. Mean arterial pressure (BP), systemic vascular resistance (SVR), and plasma histamine concentration (mean ± SE) before and after morphine 1 mg/kg and fentanyl 50 µg/kg (both infused over 10 minutes). Morphine, but not fentanyl, causes significant decrements in BP and SVR, which parallel the increase in plasma histamine concentration. (Reprinted with permission from Rosow CE, Moss J, Philbin DM, et al: Histamine release during morphine and fentanyl anesthesia. Anesthesiology 1982; 56: 93.)

Morphine does not directly affect cerebral circulation, but with morphine-induced respiratory depression, CO2 retention causes cerebral vasodilation and an elevation in cerebrospinal fluid pressure. This effect is not seen when mechanical ventilation is used to prevent hypercarbia.1 Thus, morphine and other µ agonists must be used cautiously in spontaneously breathing patients with head injury or other conditions associated with elevated intracranial pressure.

Disposition Kinetics

Morphine is rapidly absorbed after intramuscular, subcutaneous, and oral administration. Following intramuscular administration, peak plasma concentration is seen at 20 minutes and absorption half-life is estimated at 7.7 minutes (range, 2 to 15 minutes).69 After IV administration morphine undergoes rapid redistribution, with a mean redistribution half-time between 1.5 and 4.4 minutes in awake and anesthetized adults.69,70,71 Morphine has a terminal elimination half-life between 1.7 and 3.3 hours.70,71,72 Age affects morphine pharmacokinetics. The average elimination half-life of morphine is 7 to 8 hours in neonates <1 week of age and 3 to 5 hours in older infants.73 In patients between 61 and 80 years old, morphine's terminal elimination half-life was 4.5 hours compared with 2.9 hours in younger patients.72

Morphine is about 35% protein bound, mostly to albumin.17 Its steady-state volume of distribution is large, with estimates in the range of 3 to 4 L/kg in normal adults.69,70,71Morphine's major metabolic pathway is hepatic phase II conjugation, to form morphine-3-glucuronide (M3G) and morphine-6β-glucuronide (M6G). 3-Glucuronidation is the predominant pathway, and following a single IV dose, 40% and 10% of the dose are excreted in the urine as M3G and M6G, respectively.74 Unchanged morphine in the urine accounts for only about 10% of the dose. The rate of hepatic clearance of morphine is high, with a hepatic extraction ratio of 0.7.69 Thus, morphine elimination may be slowed by processes that decrease hepatic blood flow.71 Extrahepatic sites such as kidney, intestine, and lung have been suggested for morphine glucuronidation, but their importance in humans is unknown.

Active Metabolites

M6G possesses significant µ receptor affinity and potent antinociceptive activity. Appreciable plasma concentrations of M6G and M3G have been measured in cancer patients receiving high doses of oral morphine. During chronic oral morphine therapy, plasma M6G concentrations can be higher than those of the parent morphine compound.75 Because morphine glucuronides are eliminated by the kidney, it is not surprising that very high M6G-to-morphine ratios have been reported in patients with renal dysfunction. This accumulation of the active metabolite is thought to be responsible for the unusual sensitivity of renal failure patients to morphine. While common wisdom suggests that glucuronide conjugates do not penetrate the blood–brain barrier, M6G concentration in cerebrospinal fluid is 20 to 80% that of morphine.76 Despite animal literature demonstrating the analgesic potency of M6G, there is little information in humans concerning the magnitude of analgesia and side effects of M6G relative to morphine. Portenoy et al.76 demonstrated that in cancer patients receiving chronic morphine therapy, pain relief correlated positively with the M6G-to-morphine ratio, suggesting a contributing role of M6G to overall morphine analgesia. In a study of cancer patients who received synthetic M6G (up to 60 µg/kg), 17 of 19 patients experienced effective analgesia and no adverse effects.77 In contrast, dizziness, nausea, sedation, muscle aches, and respiratory depression have been reported in volunteers who received M6G.75 However, the role of M6G in acute dosing is unclear because there is a long delay (6 to 8 hours) between the time course of plasma concentration and CNS effects.78 While the contribution of M6G to morphine-induced analgesia and side effects remains to be determined, morphine should be administered cautiously to patients with renal failure.

Dosage and Administration of Morphine

In current clinical practice morphine is used mainly as a premedicant and for postoperative analgesia, and less often as a component of balanced or high-dose opioid anesthesia. Intravenous analgesic doses of morphine for adults typically range from 0.01 to 0.20 mg/kg. When used in a balanced anesthetic technique with N2O, morphine can be given in total doses of up to 3 mg/kg with remarkable hemodynamic stability, but awareness under anesthesia is a risk. When combined with other inhalation agents, it is unlikely that more than 1 to 2 mg/kg of morphine is necessary. The morphine dose associated with apparent cardioprotective effect is a single dose of 40 mg, given before cardioplegia and cardiopulmonary bypass.68 Because of its hydrophilicity, morphine crosses the blood–brain barrier relatively slowly; and while its onset can be observed within 5 minutes, peak effects may be delayed for 10 to 40 minutes. This delay makes morphine more difficult to titrate as an anesthetic supplement than the more rapidly acting opioids.


Meperidine, a phenylpiperidine derivative (Fig. 19-5), was the first totally synthetic opioid. It was initially studied as an anticholinergic agent, but was found to have significant analgesic activity.1

Analgesia and Effect on Minimum Alveolar Concentration of Volatile Anesthetics

Meperidine's analgesic potency is about one-tenth that of morphine's and is most likely mediated by µ-opioid receptor activation. However, meperidine also has moderate affinity for κ- and δ-opioid receptors.1,79 Unlike morphine, meperidine plasma concentrations correlate reasonably well with analgesic effects.80 Although there is considerable interpatient variability, the MEAC of meperidine is approximately 200 ng/mL. There is very little information available on the effect of meperidine on the MAC of inhaled anesthetics, but a study in dogs demonstrated a dose-dependent reduction in the MAC of halothane.81

Meperidine also has well-recognized weak local anesthetic properties. Compared with morphine, fentanyl, and buprenorphine injected perineurally, only meperidine alters nerve conduction and produces analgesia.82 This has led to some popularity for epidural and subarachnoid administration, particularly in obstetric anesthesia. But because of its local anesthetic effects, neuraxial meperidine may also produce sensory and motor blockade as well as sympatholytic effects that are not seen with other opioids.

Side Effects

Like morphine, therapeutic doses of meperidine can produce sedation, pupillary constriction, and euphoria, and very high


doses are associated with CNS excitement and seizures (see later discussion). In equianalgesic doses, meperidine produces respiratory depression equal to that of morphine, as well as nausea, vomiting, and dizziness, particularly in ambulatory patients.1

Figure 19-5. Chemical structures of phenylpiperidine, meperidine, and the 4-anilinopiperidine derivatives fentanyl, sufentanil, alfentanil, and remifentanil.

Like other opioids, meperidine causes significant delay in gastric emptying. While meperidine does increase common bile duct pressure, this occurs to a lesser extent than with equianalgesic doses of morphine and fentanyl (Fig. 19-6).56,83

Analgesic doses of meperidine in awake patients are not associated with hemodynamic instability, but 1 mg/kg in patients with cardiac disease decreased heart rate, cardiac index, and rate–pressure product.84 In an isolated papillary muscle preparation, high concentrations of meperidine depressed contractility. This effect was not naloxone-reversible and is consistent with a nonspecific, local anesthetic effect.85 In higher doses, meperidine causes significantly more hemodynamic instability than morphine or fentanyl and its derivatives,86an effect at least partially related to histamine release. In a comparison of opioids administered as part of balanced anesthesia, Flacke et al.86 found that 25% patients in the meperidine group experienced severe hypotension and had abnormally elevated plasma histamine concentrations. Interestingly, only one patient in the morphine group (0.6 mg/kg morphine given) had a similar histamine plasma concentration. Thus, meperidine is not recommended in high doses for clinical anesthesia.


Meperidine is effective in reducing shivering from diverse causes, including general and epidural anesthesia, fever, hypothermia, transfusion reactions, and administration of amphotericin B. Meperidine reduces or eliminates visible shivering as well as the accompanying increase in oxygen consumption87 following general and epidural anesthesia. Equianalgesic doses of fentanyl (25 µg) and morphine (2.5 mg) did not reduce postoperative shivering, suggesting that the antishivering effect of meperidine is not mediated by µ-opioid receptors. This effect may be mediated by κ-opioid receptors. Butorphanol, a drug with significant κ agonist activity, effectively reduces postoperative shivering in a dose of 1 mg.88 Furthermore, low doses of naloxone, sufficient to block µ receptors, did not reverse the antishivering effect of meperidine, but high-dose naloxone, designed to block both µ and κ receptors, did reverse the antishivering effect.79 The observation that other types of drugs, such as α1-adrenergic agonists (clonidine 1.5 µg/kg), serotonin antagonists,89 and propofol,90 can reduce postoperative shivering suggests that a nonopioid mechanism may be involved. Physostigmine 0.04 mg/kg can also prevent postoperative shivering, suggesting a role for the cholinergic system.91

Disposition Kinetics

Following IV administration, meperidine plasma concentration falls rapidly. Meperidine's redistribution half-life is 4 to 16 minutes, and its terminal elimination half-life is between 3 and 5 hours.92,93 The elimination half-life is not prolonged in elderly patients; however, in neonates and infants, the median elimination half-life is 8 to 10 hours, with greater individual variability (three- to fivefold) compared with adults.

Meperidine is moderately lipid soluble, and is 40 to 70% protein bound, mostly to albumin and α1-acid glycoprotein


(Table 19-2).94 Meperidine has a large steady-state volume of distribution, with estimates in the range of 3.5 to 5 L/kg in adults.92,93 The high clearance rate (10 mL/kg/min) reflects a high hepatic extraction ratio; it is N-demethylated in the liver to form normeperidine, the principal metabolite, and also hydrolyzed to meperidinic acid. Both metabolites may then be conjugated1 and excreted renally. Normeperidine is pharmacologically active and potentially toxic (see later discussion).

Figure 19-6. The effect of several opioids on common bile duct pressures in patients anesthetized with enflurane and N2O–O2. Patients received either fentanyl 100 µg/70 kg, morphine 10 mg/70 kg, meperidine 75 mg/70 kg, or butorphanol 2 mg/70 kg. After 20 minutes, the effects were reversed with naloxone. (Reprinted with permission from Radnay PA, Duncalf D, Novakovik M, et al: Common bile duct pressure changes after fentanyl, morphine, meperidine, butorphanol, and naloxone. Anesth Analg 1984; 63: 441.)

Active Metabolites

Normeperidine has appreciable pharmacologic activity and can produce signs of CNS excitation. Mood alterations such as apprehension and restlessness, as well as neurotoxic effects such as tremors, myoclonus, and seizures, have been reported.95 The elimination half-life of the metabolite normeperidine (14 to 21 hours) is considerably longer than the parent compound, and therefore is likely to accumulate with repeated or prolonged administration, particularly in patients with renal dysfunction.95 Myoclonus and seizures have been reported in patients receiving meperidine for postoperative or chronic pain. Patients who developed seizures had a mean plasma normeperidine concentration of 0.81 µg/mL.95 It appears that a total daily meperidine dosage of 1,000 mg is associated with an increased risk of seizures, even in patients without renal dysfunction.

Dosage and Administration of Meperidine

A single dose of meperidine is approximately one-tenth as potent as morphine when given parenterally, but has a shorter duration of action. Intravenous analgesic doses of meperidine for adults typically range from 0.1 to 1 mg/kg. Intravenous doses of 12.5 to 50 mg are effective in reducing postoperative shivering. As discussed earlier, high doses of meperidine for intraoperative use are not recommended because of hemodynamic instability. In addition, large single doses or prolonged administration may produce seizures because of the metabolite normeperidine; the total daily dose should not exceed 1,000 mg in 24 hours.


Methadone, a synthetic opioid introduced in the 1940s, is primarily a µ agonist with pharmacologic properties that are similar to morphine. Although its chemical structure is very different from that of morphine, steric factors force the molecule to simulate the pseudopiperidine ring conformation that appears to be required for opioid activity.1 Because of its long elimination half-life, methadone is most often used for long-term pain management and for treatment of opioid abstinence syndromes.

Analgesia and Use in Anesthesia

Following parenteral administration, the onset of analgesia is rapid, within 10 to 20 minutes. After single doses of up to 10 mg, the duration of analgesia is similar to morphine,1 but with large or repeated parenteral doses, prolonged analgesia can be obtained. Several investigators have administered methadone intra- and postoperatively with the aim of providing prolonged postoperative analgesia. Patients who received 20 mg methadone intraoperatively and up to 20 mg additional methadone in the immediate postoperative period had a median duration of postoperative analgesia of over 20 hours.96,97

Side Effects

Side effects of methadone are similar in magnitude and frequency to those of morphine.1,96 Patients who received 20 mg methadone at the beginning of surgery were sedated in the immediate postoperative period but did not appear to have clinically significant respiratory depression. About 50% experienced nausea or vomiting, which was easily treated with standard antiemetic therapy.96 Methadone produces typical opioid effects on smooth muscle. Like morphine, it markedly decreases intestinal propulsive activity and can cause constipation as well as biliary spasm.1

Disposition Kinetics

Following an IV dose, the plasma concentration–time data for methadone are described by a biexponential equation. The


mean redistribution half-time is 6 minutes (range, 1 to 24 minutes), and the mean terminal elimination half-time is 34 hours (range, 9 to 87 hours).97 Methadone is well absorbed after an oral dose, with bioavailability approximately 90%, and reaches peak plasma concentration at 4 hours after oral administration.1 It is nearly 90% plasma protein bound and undergoes extensive metabolism in the liver, mostly N-demethylation and cyclization to form pyrrolidines and pyrroline.1

Dosage and Administration of Methadone

The use of methadone in clinical anesthesia has focused on attempts to achieve prolonged postoperative analgesia, providing that an adequate initial dose is administered. Because adverse effects can also be prolonged, careful titration of the dose is necessary. In opioid-naïve patients, an initial single dose of 20 mg can provide analgesia without significant postoperative respiratory depression.96 Wangler and Rosenblatt98 described a technique to avoid respiratory depression in which 8 to 12 mg methadone is administered to the awake patient until the threshold of respiratory depression (respiratory rate of 6 to 8/min) is reached. Immediately prior to incision, an additional dose equal to half the initial dose is given. For administering supplemental analgesic doses in the immediate postoperative period, it is essential to confirm that patients with ongoing significant pain have no depression of respiration or level of consciousness, and that a 30- to 40-minute interval should elapse between 5-mg doses to allow full assessment of adverse effects. It may be easier and safer to use a sustained-release opioid preparation (containing oxycodone or morphine) with a shorter time to peak effect if a long-acting analgesic is desired. This is most easily accomplished by administering the oral medication preoperatively, but it is also important to note that these long-acting opioids are not currently approved for prophylaxis of postoperative pain.


Fentanyl and its analogs sufentanil and alfentanil are the most frequently used opioids in clinical anesthesia. Fentanyl, first synthesized in 1960, is structurally related to the phenylpiperidines (Fig. 19-5) and has a clinical potency ratio 50 to 100 times that of morphine. Clear plasma concentration-effect relationships have been demonstrated for fentanyl (Table 19-3). Scott et al.99 demonstrated progressive EEG changes with increasing serum fentanyl concentration (Fig. 19-7). During a brief infusion, the time lag between increasing serum fentanyl concentration and EEG slowing was 3 to 5 minutes. After the infusion was stopped, the resolution of EEG changes lagged behind decreasing serum fentanyl concentration by 10 to 20 minutes.


Fentanyl, a µ-opioid receptor agonist, produces profound dose-dependent analgesia, ventilatory depression, and sedation, and at high doses it can produce unconsciousness. In postoperative patients, the mean fentanyl dose requirement was 55.8 µg/hr, and mean MEAC in blood was 0.63 ng/mL.100 A large interpatient variability in MEAC (0.23 to 1.18 ng/mL) typical of opioids was observed, but over the 2-day study period, the MEAC for any individual patient remained relatively constant. In volunteers, a mean plasma fentanyl concentration of 1.3 ng/mL reduced experimental pain intensity ratings by 50%,46 consistent with other estimates of plasma fentanyl concentrations producing moderate-to-strong analgesia.101

Table 19-3 Plasma Concentration Ranges (ng/ml) for Various Therapeutic and Nontherapeutic Opioid Effectsa














Moderate-to-strong analgesia






50% MAC reduction







Surgical analgesia with ~70% N2O







Respiratory depression threshold






50% ↓ ventilatory response to CO2













Unconsciousness (not reliably achieved with opioids alone)





aEffects were generally achieved during continuous infusions or patient-controlled analgesia systems. Note that plasma concentrations associated with measurable depression of ventilatory drive are similar to those associated with analgesia for all opioids.
MEAC, minimum effective analgesic concentration, defined in most studies as the plasma opioid concentration associated with just perceptible analgesia; MAC, minimum alveolar concentration; NA, information not available.


Use in Anesthesia

Fentanyl reduces the MAC of volatile anesthetics in a concentration- or dose-dependent fashion. A single IV bolus dose of fentanyl 3 µg/kg, given 25 to 30 minutes prior to incision, reduced both isoflurane and desflurane MAC by approximately 50%.102 Fentanyl 1.5 µg/kg administered 5 minutes prior to skin incision reduces the minimum alveolar concentration that blocks adrenergic responses to stimuli (MAC-BAR) of isoflurane or desflurane in 60% N2O by 60 to 70%.103 No further drop is seen with an increase in fentanyl dose to 3 µg/kg. During constant plasma concentration of 0.5 to 1.7 ng/mL, fentanyl reduced isoflurane MAC by 50%.104 Fentanyl produces a steep plasma concentration-related reduction in sevoflurane MAC105; 3 ng/mL provides a 59% reduction, but a ceiling effect is reached, such that a threefold increase to 10 ng/mL reduced MAC by only an additional 17%.

Figure 19-7. The time course of electroencephalogram (EEG) spectral edge and serum concentrations of fentanyl (A) and alfentanil (B). Infusion rates were 150 µg/min fentanyl and 1,500 µg/min alfentanil. Increasing opioid effect is seen as a decrease in spectral edge. Changes in spectral edge follow serum concentrations more closely with alfentanil than with fentanyl. (Reprinted with permission from Scott JC, Ponganis KV, Stanski DR: EEG quantitation of narcotic effect: The comparative pharmacodynamics of fentanyl and alfentanil. Anesthesiology 1985; 62: 234.)

Epidural fentanyl also reduces inhaled anesthetic requirements.106 Epidural fentanyl 1, 2, and 4 µg/kg reduced halothane MAC by 45, 58, and 71%, respectively, while the same doses of fentanyl given IV reduced halothane MAC to a lesser extent, by 8, 40, and 49%, respectively.

Combining opioids with propofol rather than an inhalation agent is a technique for providing general anesthesia, referred to as total intravenous anesthesia or TIVA. For an IV anesthetic, the potency index is described as the plasma concentration required to prevent a response in 50% (CP50) or 95% (CP95) of patients to various surgical stimuli. Plasma concentrations of fentanyl and propofol that reduce hemodynamic or somatic responses to various surgical stimuli in 50% of patients have been determined using computer-assisted infusion.107 Fentanyl plasma concentrations of 1.2, 1.8, and 2.8 ng/mL were required for 50% reductions in propofol's CP50s for skin incision, peritoneal incision, and abdominal retraction, respectively. Greater fentanyl concentrations were required to suppress hemodynamic responses to these same stimuli. Thus, fentanyl reduces requirements for both volatile agents and propofol by a similar proportion.

Computer-assisted infusion of fentanyl has been included as a component of a balanced anesthetic technique.108,109 In combination with 50 to 70% N2O in oxygen, loss of consciousness, and absence of response to skin incision are achieved at plasma fentanyl concentrations of 15 to 25 ng/mL and >3.7 ng/mL, respectively. Intraoperative concentration requirements varied between 1 and 9 ng/mL. Spontaneous ventilation returned when the fentanyl concentration dropped to 1.5 to 2 ng/mL.108,109

Fentanyl has been used as the sole agent for anesthesia, a technique that requires a large initial dose of 50 to 150 µg/kg or stable plasma fentanyl concentrations in the range of 20 to 30 ng/mL.101 The major advantage of this technique is reliable hemodynamic stability. High doses of fentanyl significantly blunt the “stress response”—that is, hemodynamic and hormonal responses to surgical stimuli—while producing only minimal cardiovascular depression. Thus, the technique is sometimes referred to as stress-free anesthesia. There are also disadvantages to using high-dose fentanyl as the sole anesthetic agent. It precludes early extubation and “fast-track” techniques because of prolonged respiratory depression accompanying high-dose fentanyl. Furthermore, it appears that no dose of fentanyl will completely block hemodynamic or hormonal responses in all patients.110 Finally, there have been reports of intraoperative awareness and recall in patients who received very high doses (>50 µg/kg) of fentanyl. Because opioids do not produce muscle relaxation, and high-dose fentanyl can produce muscle rigidity, a muscle relaxant is generally required to achieve adequate surgical conditions. This can potentially increase the difficulty in detecting signs of intraoperative awareness.

Other Central Nervous System Effects

The effects of fentanyl on cerebral blood flow and intracranial pressure (ICP) have been studied in patients with and without neurologic disease. An induction dose of 16 µg/kg increased middle cerebral artery flow by 25% in normal patients having noncranial neurosurgery.111 A smaller dose (3 µg/kg) resulted in an elevation in ICP in ventilated patients with head trauma,112 but in brain tumor patients, a dose of 5 µg/kg of fentanyl with N2O–O2 did not result in elevated ICP.113 In all cases of elevation in ICP and cerebral blood flow, there were decreases in mean arterial pressure, which may have contributed to these changes.

The muscle rigidity often seen on induction with high-dose fentanyl and its derivatives may make it difficult or impossible to ventilate the patient. In a study in normal volunteers, 1,500 µg fentanyl infused over 10 minutes produced rigidity in 50% of subjects.114 A similar incidence, 35%, was seen in patients receiving 750 to 1,000 µg fentanyl during induction of general anesthesia, and up to 80% of patients receiving 30 µg/kg developed moderate-to-severe rigidity.115 Muscle rigidity seen


with high doses of fentanyl increases with age115 and is accompanied by unconsciousness and apnea,114,115 but lower doses, 7 to 8 µg/kg, have produced chest wall rigidity without unconsciousness or apnea. Streisand et al.114 hypothesized that hypercarbia from fentanyl-induced respiratory depression may have influenced fentanyl ionization and cerebral blood flow and hence the delivery of fentanyl to brain tissue. It would follow that patients instructed to deep-breathe during fentanyl induction may experience less rigidity during induction of anesthesia. This is consistent with observations by Lunn et al.116 During high-dose fentanyl induction (75 µg/kg), PaCO2 was maintained at 35 to 40 torr by assisting and then controlling respirations. Although chest wall compliance was reduced in 4 of 18 patients, no patient developed rigidity sufficient to impair ventilation.

Fentanyl has been associated with seizurelike movements during anesthetic induction, which are not associated with seizure activity on the EEG.117 Such activity may represent myoclonus, a result of opioid-mediated blockade of inhibitory motor pathways of cortical origin, or may represent exaggerations of opioid-induced muscle rigidity.117 However, fentanyl can activate epileptiform EEG activity in patients having surgery for intractable temporal lobe epilepsy.118

Fentanyl-induced pruritus typically presents as facial itching, but can be generalized. Equianalgesic plasma concentrations of fentanyl, morphine, and alfentanil produce equivalent intensity of pruritus.46 Fentanyl has also been reported to have a tussive effect. The mechanism is unclear, and it is not attenuated by pretreatment with atropine or midazolam.119

Respiratory Depression

Fentanyl produces approximately the same degree of ventilatory depression as equianalgesic doses of morphine.46 Respiratory depression—expressed as an elevation in end-tidal CO2, a decrease in the slope of the CO2 response curve, or the minute ventilation at an end-tidal CO2 of 50 mm Hg (VE50)—develops rapidly, reaching a peak in ~5 minutes,99,120,121 and the time course closely follows plasma fentanyl concentration.120,122 Even at plasma concentrations associated with mild analgesia, ventilatory depression can be detected, and the magnitude of respiratory depression is linearly related to intensity of analgesia (Table 19-3).46,123 In postoperative patients, plasma fentanyl concentrations of 1.5 to 3.0 ng/mL were associated with a 50% reduction in CO2 responsiveness.124

Fentanyl's respiratory depression is greatly increased when it is given in combination with another respiratory depressant such as midazolam. Bailey et al.121 determined that midazolam alone (0.05 mg/kg) did not depress ventilation or cause hypoxemia. Fentanyl alone (2 µg/kg) reduced the slope of the CO2 response curve and the VE50 by 50%, and 6 of 12 subjects became hypoxemic. Fentanyl and midazolam produced no greater depression of the ventilatory response to CO2 than fentanyl alone, but 11 of 12 subjects became hypoxemic and 6 of 12 became apneic within 5 minutes. These observations suggest that this frequently used combination blunts the hypoxic ventilatory drive to a greater extent than the hypercarbic ventilatory drive. Precautions such as supplemental oxygen and pulse oximetry monitoring are recommended when such drug combinations are used.

Airway Reflexes

Although obtundation of airway reflexes by general inhalation anesthetics is well described, little is known about the direct effects of opioids on these protective reflexes. Tagaito et al.125 examined the dose-related effects of fentanyl on airway responses to laryngeal irritation during propofol anesthesia in humans. All patients had laryngeal mask airways; half breathed spontaneously and half had ventilation controlled to maintain an end-tidal CO2 of 38 mm Hg. In both groups, stimulation of the larynx (application of water to mucosa) elicited a forced expiration, followed by spasmodic panting mingled with cough reflexes and brief laryngospasm. With three cumulative fentanyl doses (50, 50, and 100 µg), expiration, panting, and coughing decreased in a dose-dependent fashion. Apnea with laryngospasm occurred after the first dose, but with cumulative fentanyl dosing, the duration of laryngospasm shortened. Cough was the airway reflex most vulnerable to depression by fentanyl. Attenuation of airway reflexes is desirable during general anesthesia, but it is equally desirable that these protective reflexes return to baseline rapidly after emergence, and remain intact throughout conscious sedation. Doses required to suppress cough and other reflexes in awake or sedated individuals have not been characterized.

Cardiovascular and Endocrine Effects

Isolated heart muscle models demonstrate concentration-dependent negative inotropic effects of opioids, including morphine, meperidine, and fentanyl.65 A very high fentanyl concentration (10 µg/mL) reduced contractility by 50%, but 1 µg/mL had no significant effects on papillary muscle mechanics. In clinical practice, even high-dose fentanyl administration (up to 75 µg/kg) produces much lower plasma concentrations, in the range of 50 ng/mL,116 and is associated with remarkable hemodynamic stability. Patients who received 7 µg/kg fentanyl at induction of anesthesia had a slight decrease in heart rate, but no change in mean arterial pressure compared with control.86 Fentanyl-induced bradycardia is more marked in anesthetized than conscious subjects, and usually resolves with atropine. With higher fentanyl doses, in the range of 20 to 25 µg/kg, decreases in heart rate, mean arterial pressure, systemic and pulmonary vascular resistance, and pulmonary capillary wedge pressure of approximately 15% were seen in patients with coronary artery disease.116,126 Very high fentanyl doses, up to 75 µg/kg, produced no further hemodynamic changes. All of these patients had been premedicated with diazepam, pentobarbital, scopolamine, and/or atropine. In unpremedicated patients undergoing noncardiac surgery, induction with fentanyl 30 µg/kg produced no changes in heart rate or systolic blood pressure.115 Hypertension in response to sternotomy is the most common hemodynamic disturbance during high-dose fentanyl anesthesia and occurs in 40 and 100% in patients receiving 50 to 100 µg/kg.127 Unlike morphine and meperidine, which induce hypotension, at least in part because of histamine release,64,128 high-dose fentanyl (50 µg/kg) is not associated with significant histamine release (Fig. 19-4).

Although high doses of fentanyl are associated with minimal cardiovascular changes, combining fentanyl with other drugs can compromise hemodynamic stability. The combination of fentanyl and diazepam produces significant cardiovascular depression.115,126 Diazepam 10 mg given after 20 to 50 µg/kg of fentanyl decreased stroke volume, cardiac output, systemic vascular resistance, and mean arterial pressure, and increased central venous pressure significantly.126 Adding 60% N2O to high-dose fentanyl produced a significant decrease in cardiac output and increases in systemic and pulmonary vascular resistance.116

High-dose fentanyl (100 µg/kg) prevented increases in plasma epinephrine, cortisol, glucose, free fatty acids, and growth hormone (the “stress response”) during surgery, but a lower dose of fentanyl (5 µg/kg followed by an infusion of 3 µg/kg/h) did not.129 Unlike morphine, fentanyl does not prevent the inflammatory effects associated with cardiopulmonary bypass, nor does it produce the apparent cardioprotective effects seen with morphine.67,68


Smooth Muscle and Gastrointestinal Effects

Fentanyl, like morphine and meperidine, significantly increases common bile duct pressure (Fig. 19-6).83 Like other opioids, fentanyl can cause nausea and vomiting, particularly in ambulatory patients, and can delay gastric emptying and intestinal transit.

Disposition Kinetics

Fentanyl's extreme lipid solubility (Table 19-2) allows rapid crossing of biologic membranes and uptake by highly perfused tissue groups, including the brain, heart, and lung. Thus, after a single bolus dose, the onset of effects is rapid and the duration brief. Hug and Murphy130 determined the relationships between fentanyl effects and its concentration over time in plasma and various tissues in rats given fentanyl 50 µg/kg (Fig. 19-8). The onset of opioid effects occurred within 10 seconds and correlated with a rapid increase in brain tissue fentanyl concentration, which equilibrated with plasma by 1.5 minutes. Recovery from fentanyl effects started within 5 minutes and was complete by 60 minutes. Elimination from the “central tissues” (brain, heart, and lung) was also rapid, as fentanyl was redistributed to other tissues, particularly muscle and fat. Peak muscle concentration was seen at 5 minutes, while fat concentration reached a maximum approximately 30 minutes after the dose. The delay in fat uptake despite fentanyl's high lipid solubility is because of the limited blood supply to that tissue. Thus, redistribution to muscle and fat limits the duration of a bolus dose of fentanyl, and accumulation in peripheral tissue compartments can be extensive because of the large mass of muscle and high affinity of fentanyl for fat. With prolonged administration of fentanyl, fat can act as a reservoir of drug.

Fentanyl pharmacokinetics has been studied in awake and anesthetized individuals. After an IV dose, plasma fentanyl concentration falls rapidly, and the concentration-time curve has been described by both two- and three-compartment models.130 McClain and Hug120 administered fentanyl 3.2 or 6.4 µg/kg to healthy male volunteers and found that nearly 99% of the dose was eliminated from plasma by 60 minutes. These investigators found both rapid and slower distribution phases, with half-times of 1.2 to 1.9 minutes and 9.2 to 19 minutes, respectively. The terminal elimination half-time ranged from 3.1 to 6.6 hours, somewhat longer than that for morphine. Similar values were noted in surgical patients <50 years old,113,131 including morbidly obese patients.113 Reports of age effects on fentanyl kinetics are conflicting. The fentanyl requirement decreases with increasing age (20 to 89 years), but pharmacokinetic parameters do not change.132 In contrast, Bentley et al.131 observed a marked decrease in clearance and an increase in terminal elimination half-time to approximately 15 hours in patients >60 years old compared with 4.4 hours in patients <50 years old.

Figure 19-8. Fentanyl uptake and elimination in various tissues of the rat following intravenous injection. Unchanged fentanyl tissue concentrations (means for six rats) are expressed as percentage of dose. “Central” represents the combined content of brain, heart, and lung tissues. The large mass of muscle (50% body weight of the rat) and high affinity of fat for fentanyl (despite slow equilibration) serve as a drain on the central compartment. (Reprinted with permission from Hug CC, Murphy MR: Tissue redistribution of fentanyl in terms of its effects in rats. Anesthesiology 1981; 55: 369.)

Unlike its derivatives, fentanyl is significantly bound to red blood cells, approximately 40%, and has a blood: plasma partition coefficient of approximately 1.133 Plasma fentanyl is highly protein bound, with estimates in the range of 79 to 87%. It binds avidly to α1-acid glycoprotein but also binds to albumin.133,134 Fentanyl protein binding is pH-dependent, such that a decrease in pH will increase the proportion of fentanyl that is unbound.133 Thus, a patient with respiratory acidosis will have a higher proportion of unbound (active) fentanyl, which could exacerbate respiratory depression. Clearance of fentanyl is primarily by rapid and extensive metabolism in the liver. Clearance estimates of 8 to 21 mL/kg/min approach liver blood flow and indicate a high hepatic extraction ratio.120,132 Thus, hepatic metabolism of fentanyl is expected to be dependent on liver blood flow. Metabolism is primarily byN-dealkylation to norfentanyl and by hydroxylation of both the parent and norfentanyl.133 Only about 6% of the dose of fentanyl is excreted unchanged in the urine.120

Dosage and Administration of Fentanyl

From administration as a single bolus dose, fentanyl developed an early reputation as a short-acting opioid, but experience with very large doses and multiple doses revealed that prolonged respiratory depression and delayed recovery could occur (Table 19-4). These observations demonstrate that fentanyl's clinical duration is limited by redistribution, and that with prolonged administration, accumulation can occur, as discussed later in this chapter.

Fentanyl can be useful as a sedative/analgesic premedication when given a short time prior to induction. For this use, incremental doses of 25 to 50 µg IV are titrated until the desired effect is achieved. It is important to note that although the onset of fentanyl's effects is rapid, peak effect lags behind peak plasma concentration by up to 5 minutes.99 A transmucosal delivery system for fentanyl is also available and has been shown to be an effective premedicant for pediatric and adult patients as well as an effective treatment for “breakthrough” pain in chronic pain patients. Doses of 10 to 20 µg/kg in children and 400 to 800 µg in adults, administered 30 minutes prior to induction or a painful procedure, are safe and effective, but dose-dependent side effects typical of opioids are reported.135,136 Because respiratory depression and hypoxemia can occur, transmucosal fentanyl usually should be administered in a monitored environment.

Fentanyl is used frequently as an adjunct to induction agents to blunt the hemodynamic response to laryngoscopy and tracheal intubation, which can be particularly severe in patients with hypertension or cardiovascular disease. Common clinical practice involves titration of fentanyl in doses of 1.5 to 5 µg/kg prior to administration of the induction agent. Because its peak effect lags behind peak plasma concentration by 3 to 5 minutes, fentanyl titration should be complete approximately 3 minutes prior to laryngoscopy to maximally blunt hemodynamic responses to tracheal intubation. Perhaps the most common clinical use of fentanyl and its derivatives is


as an analgesic component of balanced general anesthesia. With this technique, incremental doses of fentanyl 0.5 to 2.5 µg/kg are administered intermittently as dictated by the intensity of the surgical stimulus and may be repeated approximately every 30 minutes. Generally, administration of up to 3 to 5 µg/kg/hr will allow recovery of spontaneous ventilation at the end of surgery. As an alternative to intermittent dosing, a loading dose of 5 to 10 µg/kg and continuous fentanyl infusion at a rate between 2 and 10 µg/kg/hr are recommended.109 It is important to remember, however, that anesthetic requirements vary with age, concurrent diseases, and the surgical procedure. For example, fentanyl requirements decrease by 50% as age increases from 20 to 89 years.132 Fentanyl requirements can also be expected to decrease with the duration of infusion (see “Context-Sensitive Half-Time”).

Table 19-4 Dosage for Fentanyl, Sufentanil, Alfentanil, and Remifentanil During Elective Surgery in Adultsa






Premedication (µg)






   With hypnotic (µg/kg)




and/or 0.25–0.5 µg/kg/min

   With 60–70% N2O (µg/kg)




   High-dose opioid (µg/kg)




+/or 2 µg/kg/min


   Balanced anesthesia

      Intermittent bolus (µg)





      Infusion (µg/kg/min)





   High dose opioid (µg/kg/min)




Transition to PACU (µg/kg/min)



Monitored Anesthesia Care

   Intermittent bolus (µg)





   Infusion (µg/kg/min)


PACU, postanesthesia care unit.

aDoses are guidelines for hemodynamically stable adults. They should be adjusted downward for elderly patients and those with cardiac dysfunction and hemodynamic instability.

Fentanyl combined with high-dose droperidol and nitrous oxide, a technique called neuroleptanesthesia,137 is rarely used today because of concerns about prolongation of the QT interval of the electrocardiogram by high-dose droperidol.138 High-dose (e.g., 50 to 150 µg/kg) fentanyl “anesthesia” has been used extensively for cardiac surgery. With this technique, a mean plasma fentanyl concentration of 15 ng/mL, which prevents hemodynamic changes in response to noxious stimuli,139 can be achieved with a loading dose of 50 µg/kg, followed by a continuous infusion of 30 µg/kg/hr. With high-dose fentanyl, muscle relaxants and mechanical ventilation are required.

The use of fentanyl in the management of acute and chronic pain is discussed in Chapters 57 and 58).


Sufentanil, a thienyl derivative of fentanyl (Fig. 19-5) first described in the mid-1970s, has a clinical potency ratio 2,000 to 4,000 times that of morphine and 10 to 15 times that of fentanyl.140,141 Like fentanyl, sufentanil equilibrates rapidly between blood and brain, and demonstrates clear plasma concentration-effect relationships. In a study comparing effects of sufentanil and fentanyl on the EEG, Scott et al.141 noted similar pharmacodynamic profiles. During a 4-minute sufentanil infusion, the change in spectral edge lagged behind the rising sufentanil concentration by approximately 2 to 3 minutes, while resolution of the EEG changes lagged behind plasma concentration changes by 20 to 30 minutes.


Sufentanil is a highly selective µ-opioid receptor agonist and exerts potent analgesic effects in animals when given by either systemic or spinal routes. While the literature describing clinical experience with sufentanil as a component of general anesthesia is extensive, available information regarding the analgesic potency of systemically administered sufentanil in humans is limited. Geller et al.142 titrated an IV infusion rate to adequate postoperative analgesia, and noted that a mean rate of 8 to 17 µg/hr was required during the first 48 hours. This was associated with a fivefold range in plasma sufentanil concentrations, between 0.02 and 0.1 ng/mL. Similar sufentanil requirements (including a wide interpatient variability) are noted in other studies of postoperative and cancer patients.47,143 Lehman et al.143 estimated that the MEAC of sufentanil is near 0.03 ng/mL, and the analgesic EC50 of sufentanil concentration is approximately 0.05 ng/mL.144

Use in Anesthesia

In animal studies, sufentanil decreases the MAC of volatile anesthetics in a dose-dependent manner, with the maximum MAC reduction between 70 and 90%.145 In humans, a plasma sufentanil concentration of 0.145 ng/mL is associated with a 50% reduction in isoflurane MAC.146 Increasing the plasma sufentanil concentration to 0.5 ng/mL reduced isoflurane MAC by 78%, and a ceiling effect was approached with greater plasma sufentanil concentrations. The maximum MAC reduction seen in humans was 89% at a sufentanil concentration of 1.4 ng/mL.

In clinical anesthesia practice, sufentanil is used as a component of balanced anesthesia and has been employed extensively in high doses (10 to 30 µg/kg) with oxygen and muscle relaxants for cardiac surgery. In this dose range, sufentanil is


at least as effective as fentanyl in its ability to produce and maintain hypnosis. In addition, hemodynamic stability appears to be as good as or better than that achieved with fentanyl.86,140 Bailey et al.147 used a computer-assisted continuous infusion system to determine the sufentanil plasma concentration response to various noxious stimuli during high-dose sufentanil anesthesia for cardiac surgery. They estimated the plasma concentration associated with a 50% probability of no response (movement, hemodynamic, or sympathetic) to intubation, incision, sternotomy, and mediastinal dissection (CP50). The CP50 for intubation, incision, and sternotomy (pooled data) was 7.06 ng/mL, and for mediastinal dissection CP50 was 12.1 ng/mL. As is typical of opioids, a wide intersubject variability (three- to tenfold) was noted in sufentanil concentration requirements. However, when used as the sole anesthetic agent, even high doses may not completely block the hemodynamic responses to noxious stimuli.110

Other Central Nervous System Effects

Equianalgesic doses of sufentanil and fentanyl produce similar changes in the EEG.140,141 In patients who received sufentanil 15 µg/kg, α activity became prominent within a few seconds, and within 3 minutes, the EEG consisted almost entirely of slow δ activity.140 Rigidity and myoclonic activity resembling seizures have been reported during induction of, and on emergence from, anesthesia with sufentanil in doses of approximately 1 to 2 µg/kg.43,44

In patients with intracranial tumors, sufentanil 1 µg/kg was associated with an elevation in spinal cerebrospinal pressure and a decrease in cerebral perfusion pressure.148 As seen with fentanyl, mean arterial pressure had dropped significantly in these patients. In normal volunteers, a smaller dose of sufentanil (0.5 µg/kg) was not associated with changes in cerebral blood flow.149 Very large doses of sufentanil (20 µg/kg) in dogs decreased cerebral blood flow in proportion to cerebral metabolism, and intracranial pressure did not change.150

Respiratory Depression

Like other µ opioid agonists, sufentanil causes respiratory depression in doses associated with clinical analgesia.122,123 Respiratory depression can be especially marked in the presence of inhalation anesthetics. In spontaneously breathing patients anesthetized with 1.5% halothane and N2O, a small dose of sufentanil (approximately 2.5 µg) reduced mean minute ventilation by 50%, and 4 µg reduced mean respiratory rate by 50%.151 Postoperative respiratory depression after apparent recovery from anesthesia has been reported for both fentanyl and sufentanil.152 The lack of exogenous stimulation in the early postoperative period may be an important factor during early recovery from anesthesia.

In normal volunteers who received bolus doses of fentanyl and sufentanil, changes in end-tidal CO2 were the same for fentanyl and sufentanil, but the slope of the ventilatory response to CO2 was depressed to a greater extent by fentanyl.122 In another volunteer study, a fourfold range of equianalgesic plasma concentrations of morphine and sufentanil produced equivalent respiratory depression, measured as both increased end-tidal CO2 and a decreased ventilatory response to CO2.144

Cardiovascular and Endocrine Effects

In animal studies, sufentanil produces vasodilation by a sympatholytic mechanism but may also have a direct smooth muscle effect.153 Clinically, a prominent feature of many trials involving sufentanil is the remarkable hemodynamic stability achieved during balanced and high-dose (up to 30 µg/kg) opioid anesthesia. Only a modest decrease in mean arterial pressure is observed when sufentanil (approximately 15 µg/kg) is used for induction of anesthesia.86,154

In general, sufentanil and fentanyl have been found to be equivalent for use in balanced and high-dose opioid anesthesia,110,155 but one clinical comparison noted better analgesia and respiratory function with sufentanil in the immediate postoperative period.156 The choice of premedication and muscle relaxant may significantly affect hemodynamics during induction and maintenance of anesthesia with sufentanil. Combining vecuronium and sufentanil can cause a decrease in mean arterial pressure during induction,157 and significant bradycardia and sinus arrest158 have been reported. Bradycardia is not seen when pancuronium is used during anesthesia with sufentanil.

Sufentanil, like fentanyl, reduces the endocrine and metabolic responses to surgery.140 However, even a large induction dose (20 µg/kg) did not prevent increases in cortisol, catecholamines, glucose, and free fatty acids during and after cardiopulmonary bypass.159

Disposition Kinetics

Sufentanil is extremely lipophilic and has pharmacokinetic properties similar to that of fentanyl. Because of a smaller degree of ionization at physiologic pH and higher degree of plasma protein binding, its volume of distribution is somewhat smaller and its elimination half-life shorter than that of fentanyl (Table 19-2). Sufentanil pharmacokinetics has been studied in anesthetized patients who had received methohexital for anesthetic induction, followed by the sufentanil dose of 5 µg/kg, and N2O in oxygen 33%.160 Plasma sufentanil concentration drops very rapidly after an IV bolus dose, and 98% of the drug is cleared from plasma within 30 minutes. Plasma concentration–time data in this study were best fitted to a three-compartment model, with rapid and slower distribution half-times of 1.4 and 17.7 minutes, respectively, and an elimination half-life of 2.7 hours. In other pharmacokinetic studies with anesthetized patients, reported mean elimination half-lives were in the range of 2.2 to 4.6 hours.161,163 Obese patients have a larger total volume of distribution and a longer elimination half-life (3.5 vs. 2.2 hours) compared with nonobese patients.161

Sufentanil is less red cell bound than fentanyl (22 compared with 40%).133 Plasma sufentanil is approximately 92% protein bound at pH 7.4, mostly to α1-acid glycoprotein. Clearance of sufentanil is rapid, and like fentanyl, it has a high hepatic extraction ratio.133 Metabolism in the liver is by N-dealkylation and O-demethylation, but sufentanil clearance and elimination half-life in patients with cirrhosis are similar to controls.162

Dosage and Administration of Sufentanil

Sufentanil is most often used as a component of balanced anesthesia, or as a single agent in high doses, particularly for cardiac surgery (Table 19-4). Several investigations have found similar sufentanil dose requirements for induction of anesthesia.86,147,163 When sufentanil is titrated during induction, loss of consciousness is seen with total doses between 1.3 and 2.8 µg/kg. Doses in the range of 0.3 to 1.0 µg/kg given 1 to 3 minutes prior to laryngoscopy can be expected to blunt hemodynamic responses to intubation, but muscle rigidity can occur even at these lower doses, particularly in the elderly.

Balanced anesthesia is maintained with intermittent bolus doses or a continuous infusion. With bolus doses of 0.1 to 0.5 µg/kg, mean maintenance requirements of 0.35 µg/kg/hr have been reported.86 Cork et al.164 administered an initial bolus of 0.5 µg/kg followed by an infusion of 0.5 µg/kg/hr, titrated to patient need. This regimen of sufentanil in combination with


N2O 70% in oxygen, with or without isoflurane, provided satisfactory anesthesia with good hemodynamic stability. Thus, for balanced anesthesia, dose requirements for bolus administration and continuous infusion are similar, in the range of 0.3 to 1 µg/kg/hr. Much higher bolus doses (10 µg/kg) and/or infusion rates (0.15 µg/kg/min) are required to achieve the plasma sufentanil concentration range of 6 to 60 ng/mL required during cardiac anesthesia using sufentanil as the sole agent.


Alfentanil, a tetrazole derivative of fentanyl (Fig. 19-5), was synthesized 2 years after sufentanil and introduced into clinical practice in the early 1980s. On a milligram basis, its clinical potency is approximately 10 times that of morphine and one-fourth to one-tenth that of fentanyl when given in single doses. Alfentanil differs from fentanyl in its pharmacokinetics as well as in its speed of equilibration between plasma and effect site in the brain. In a comparison using EEG spectral-edge effects to quantify fentanyl and alfentanil pharmacodynamics, Scott et al.99 demonstrated that alfentanil's effect followed serum drug concentration more closely than fentanyl (Fig. 19-8). Peak effect lagged behind peak plasma concentration by <1 minute, and resolution of effect followed decreasing serum alfentanil concentration by no more than 10 minutes. Alfentanil is a µ-opioid receptor agonist and produces typical naloxone-reversible analgesia and side effects such as sedation, nausea, and respiratory depression.


Clear concentration and dose-related analgesic effects have been demonstrated for alfentanil, but, as is typical for opioids, individual requirements vary widely. For postoperative analgesia, the MEAC is approximately 10 ng/mL, with a range of 2 to >40 ng/mL.165 In a laboratory investigation, 80 ng/mL was associated with a 50% reduction in pain intensity.46Similar results are seen in clinical studies, in which mean alfentanil plasma concentrations required for relief of moderate-to-severe pain are approximately 40 to 80 ng/mL (Table 19-3).166 Following an adequate loading dose, average alfentanil requirements for postoperative analgesia are approximately 10 to 20 µg/kg/hr.167,168

Use in Anesthesia

Like other opioids, alfentanil decreases the MAC of enflurane in a curvilinear fashion up to a plateau.26,169 In dogs, an infusion rate of 8 µg/kg/min (plasma concentration, 223 ng/mL) reduced enflurane MAC by 69%, but increasing the infusion rate fourfold did not reduce enflurane MAC further.158

Alfentanil plasma concentrations required to supplement N2O anesthesia have been determined.170 Patients received a loading dose of 150 µg/kg, followed by an infusion titrated between 25 and 150 µg/kg/hr according to responses to surgical stimuli. Plasma concentrations required along with 66% N2O to obtund somatic, autonomic, and hemodynamic responses to stimuli in 50% of patients were 475, 279, and 150 ng/mL for tracheal intubation, skin incision, and skin closure, respectively. The plasma alfentanil concentration associated with spontaneous ventilation after discontinuation of N2O was 223 ng/mL. Nearly identical results were obtained in a similar study using computer-controlled infusions to deliver alfentanil (Fig. 19-9).171 Plasma alfentanil concentrations required in combination with propofol to obtund responses to intubation and surgical stimuli have also been determined.172 In contrast to combining alfentanil and N2O, much lower alfentanil plasma concentrations (55 to 92 ng/mL) were required to prevent responses in 50% of patients when alfentanil was combined with propofol at a plasma concentration of 3 µg/mL (Fig. 19-10).

Figure 19-9. The relationship between alfentanil plasma concentration (with 66% N2O) and the probability of no response for intubation, skin incision, and skin closure; and the relationship of plasma alfentanil concentration (without N2O) and the recovery of adequate spontaneous ventilation. (Reprinted with permission from Ausems ME, Vuyk J, Hug CC, et al: Comparison of a computer-assisted infusion vs. intermittent bolus administration of alfentanil as a supplement to nitrous oxide for lower abdominal surgery. Anesthesiology 1988; 68: 851.)

High-dose alfentanil has been used as an induction agent for patients with and without cardiac disease173 and for induction and maintenance of cardiac anesthesia.127,174 Patients with cardiac valvular or coronary artery disease required half as much alfentanil to induce unconsciousness.173 When used as the sole anesthetic agent, mean plasma alfentanil concentrations required to significantly blunt hemodynamic responses to intubation and sternotomy were 700 to 830 ng/mL and 1,200 to 1,800 ng/mL, respectively.175 These values are approximately twice those reported for alfentanil in combination with 66% nitrous oxide.170,171 However, even doses that produced very high plasma alfentanil concentrations (1,200 to >2,000 ng/mL) did not eliminate responses to intubation and intraoperative stimuli in all patients.170 In contrast to fentanyl and sufentanil, the duration of even very large doses of alfentanil is short, so repeated doses or a continuous infusion of alfentanil is required.

Other Central Nervous System Effects

Alfentanil produces the typical generalized slowing of the EEG.99,176 Like fentanyl, alfentanil can increase epileptiform EEG activity in patients with intractable temporal lope epilepsy having surgery under general anesthesia.143 Like fentanyl and sufentanil, alfentanil can produce intense muscle rigidity accompanied by loss of consciousness. In 90 to 100% of patients, induction doses of 150 to 175 µg/kg were associated with muscle rigidity, which was not limited to the chest wall or trunk. Rather, electromyography has shown increased activity of comparable magnitude in muscles of the neck, extremities, chest wall, and abdomen.148,177

Alfentanil has been reported to increase cerebrospinal fluid pressure in patients with brain tumors, whereas fentanyl does not.113 However, when normocapnia and blood pressure were maintained at baseline, no clinically significant changes in ICP and no evidence of cerebral vasodilation or vasoconstriction were seen in neurosurgical patients who received 25 and


50 µg/kg of alfentanil with N2O.178 When the effects of three-dose regimens of alfentanil, 10, 20, and 30 µg/kg, followed by 10, 20 and 30 µg/kg/hr, were compared with placebo in brain tumor patients anesthetized with propofol and fentanyl, mean arterial pressure and cerebral perfusion pressure decreased in a dose-dependent fashion, but there were no changes in subdural ICP or arteriovenous O2 content difference.179

Figure 19-10. The alfentanil plasma concentration-effect relationships for intubation, skin incision, and the opening of the peritoneum when given as a supplement to propofol. (Reprinted with permission from Vuyk J, Lim T, Engbers FHM, et al: Pharmacodynamics of alfentanil as a supplement to propofol or nitrous oxide for lower abdominal surgery in female patients. Anesthesiology 1993; 78: 1036.)

Respiratory Depression

In animal and human studies, antinociceptive effects could not be separated from respiratory depression in volunteers; mild ventilatory depression (increased end-tidal CO2; decreased slope of the CO2 response curve) was seen at plasma concentrations as low as 20 ng/mL. At plasma concentrations associated with 50% reduction in pain intensity, respiratory depression was equivalent for alfentanil, fentanyl, and morphine.15 A clinical study examined postoperative analgesia and respiratory effects of alfentanil administered by a patient-controlled analgesia system.180 In patients who received a continuous alfentanil infusion at 900 µg/hr plus 100- to 200-µg doses as needed, three of ten patients developed respiratory depression (respiratory rate <8/min). Mean alfentanil blood concentration in this group of patients was 80 ng/mL.

Two clinical studies examined the intensity and duration of respiratory depressant effects of alfentanil in the immediate postoperative period.181,182 Patients received balanced anesthesia 67% N2O with or without 0.5% halothane and alfentanil 20 to 100 µg/kg/hr. At the end of surgery the infusion was decreased to 20 µg/kg/hr, which produced plasma alfentanil concentrations between 106 and 120 ng/mL, and good analgesia. Ventilatory response to CO2 was decreased to 50% of the baseline value, but PaCO2 was only moderately elevated (42 to 48 torr). By 2 hours after alfentanil was discontinued, respiratory function was near baseline. Recovery of ventilatory function was faster with alfentanil compared with fentanyl.182 Another comparison found that for anesthetics of 1.5 to 2 hours' duration, recovery of respiratory function was similar with alfentanil and fentanyl.183 Like its congeners, alfentanil has been associated with apnea and unconsciousness after apparent recovery from anesthesia.184

Cardiovascular Effects

The cardiovascular effects of alfentanil are influenced by preoperative medication, muscle relaxant used, method of administration, and the degree of surgical stimulation. In general, heart rate and mean arterial pressure are unchanged or slightly decreased during induction with alfentanil 40 to 120 µg/kg,173 but rapid induction with 150 to 175 µg/kg alfentanil can decrease mean arterial pressure by 15 to 20 torr. After induction with etomidate, alfentanil 120 µg/kg decreased mean arterial pressure by approximately 30 torr,185and following thiopental (3 to 5 mg/kg) induction, a smaller dose of alfentanil (40 µg/kg) decreased mean arterial pressure by approximately 40 torr.186 Alfentanil does not appear to have negative inotropic effects,185 but severe hypotension has been observed when alfentanil is given after 0.125 mg/kg diazepam.187 In combination with lorazepam premedication or thiopental induction, moderate doses (10 to 50 µg/kg) of alfentanil blunt the cardiovascular and catecholamine responses to laryngoscopy and intubation,175,186 but for patients >70 years old, doses in this range given with thiopental can produce significant hypotension after induction.188 Alfentanil can also cause bradycardia, but this effect is minimized by premedication with atropine and by the vagolytic effect of pancuronium. Alfentanil 50 µg/kg combined with propofol 1 mg/kg for induction of anesthesia can produce significant bradycardia and hypotension after intubation, but premedication with glycopyrrolate prevents these effects.189


Nausea and Vomiting

Clinical comparisons between alfentanil and sufentanil190 or fentanyl191 and N2O revealed the same incidence of nausea and vomiting. In normal volunteers receiving computer-controlled opioid infusions, the severity of nausea at equianalgesic plasma concentrations was equivalent for alfentanil, fentanyl, and morphine,46 but alfentanil-induced nausea and vomiting resolved more quickly (B.A. Coda, unpublished observations, 1988–90).

Disposition Kinetics

Alfentanil pharmacokinetics differs from fentanyl and sufentanil in several respects (Table 19-3). A unique characteristic is that alfentanil is a weaker base than other opioids. Whereas other opioids have pKa above 7.4, the pKa of alfentanil is 6.8; consequently, nearly 90% of unbound plasma alfentanil is nonionized at pH 7.4.131 This property, together with its moderate lipid solubility, enables alfentanil to cross the blood–brain barrier rapidly and accounts for its rapid onset of action. Compared with fentanyl and sufentanil, which have mean plasma-brain equilibration half-times of 6.4 and 6.2 minutes, respectively,99,141 alfentanil has a blood–brain equilibration half-time of 1.1 minutes.121 Alfentanil also has a smaller volume of distribution than fentanyl, which is a result of lower lipid solubility and high protein binding.192 Approximately 92% of alfentanil is protein bound, mostly to α1-acid glycoprotein.131,134

After IV administration, plasma alfentanil concentration falls rapidly; 90% of the administered dose has left the plasma by 30 minutes,193 mostly because of distribution to highly perfused tissues. Plasma concentration decay curves in patients most often fit a three-compartment model.24,193 Like fentanyl, alfentanil is quickly distributed, with rapid and slow distribution half-times of 1.0 to 3.5 minutes and 9.5 to 17 minutes, respectively. However, alfentanil has a terminal elimination half-life of 84 to 90 minutes, which is considerably shorter than those of fentanyl and sufentanil. Clearance of alfentanil, 6.4 mL/kg/min, is just half that of fentanyl, but because alfentanil's volume of distribution is 4 times smaller than fentanyl's, relatively more of the dose is available to the liver for metabolism.194 Chauvin et al.195 found that alfentanil has an intermediate hepatic extraction coefficient (32 to 53%) in humans, and that its elimination depends on hepatic plasma flow.

In animals, alfentanil undergoes N-dealkylation and O-demethylation in the liver to form inactive metabolites.131 Liver disease can significantly prolong the elimination half-life of alfentanil. Patients with moderate hepatic insufficiency as a result of cirrhosis have reduced binding to α1-acid glycoprotein and a plasma clearance one-half that of control patients. These changes result in a marked increase in the elimination half-life, 219 minutes versus 90 minutes in controls.196 Renal disease also decreases alfentanil protein binding, but does not result in decreased plasma clearance or a prolonged terminal elimination half-life.197 Alfentanil's elimination half-life is prolonged by about 30% in the elderly and appears to be much shorter (about 40 minutes) in children 5 to 8 years old.198 Obesity is also associated with a 50% decrease in alfentanil clearance and a prolonged (172 minutes) elimination half-life.198

The combination of moderate lipid solubility and short elimination half-life suggests that both redistribution and elimination are important in the termination of alfentanil's effects.182After a single bolus dose, redistribution will be the most important mechanism, but after a very large dose, repeated small doses, or a continuous infusion, elimination will be a more important determinant of the duration of alfentanil's effects.

Dosage and Administration of Alfentanil

Because of its rapid onset, alfentanil has been used as an induction agent alone or in combination with other drugs (Table 19-4). In healthy patients, doses of about 120 µg/kg produce unconsciousness in 2 to 2.5 minutes, but may also produce muscle rigidity. Premedication with a benzodiazepine (e.g., lorazepam 0.08 mg/kg) is associated with a lower dose requirement, 40 to 50 µg/kg, and a faster onset of unconsciousness, within 1.5 minutes,173 but may also produce hypotension. For rapid sequence induction, a bolus dose of 36 µg/kg followed by thiopental and rocuronium can yield ideal intubating conditions within 40 seconds in 95% of patients.199 A lower dose, 15 µg/kg (range, 13 to 31 µg/kg) with sevoflurane and N2O, but without muscle relaxants produced good intubating conditions within 90 seconds in 95% of patients.200 With propofol 2.5 mg/kg, an alfentanil dose of 10 µg/kg appears optimal for laryngeal mask insertion, but is accompanied by apnea for about 2 minutes.201

Because of its brief duration of action, alfentanil can be a useful component of general anesthesia in short surgical procedures, especially those associated with minimal postoperative pain, particularly in the outpatient surgery. In this setting, loading doses of 5 to 10 µg/kg provide good analgesia with rapid recovery.192 For longer procedures, alfentanil can be administered as needed in repeated small bolus doses, but its pharmacokinetic properties make it ideal for administration as a continuous infusion. After induction of anesthesia, a loading dose of alfentanil 10 to 50 µg/kg is followed with supplemental bolus doses of 3 to 5 µg/kg as needed or a continuous infusion starting at 0.4 to 1.7 µg/kg/min with 60 to 70% N2O or a propofol infusion.170,171,192,200,201,202,203 A pediatric study reported use of similar doses of alfentanil and propofol,204 while another used higher alfentanil doses (100 µg/kg loading dose followed by 2.5 µg/kg/min) combined with 70% N2O without propofol.205

When high-dose alfentanil is used as the sole anesthetic agent, a continuous infusion of up to 150 to 600 µg/kg/hr is adjusted according to the patient's responses to stimuli, but much lower doses can be effective for cardiac surgery if adequate premedication is given.174


Remifentanil, a 4-anilidopiperidine with a methyl ester side chain (Fig. 19-5) that was first described in 1990 and approved for clinical use in 1996, was developed to meet the need for an ultrashort-acting opioid. Because its ester side chain is susceptible to metabolism by blood and tissue esterases, remifentanil is rapidly metabolized to a substantially less active compound. Thus, because its ultrashort action is due to metabolism rather than to redistribution, it does not accumulate with repeated dosing or prolonged infusion. Remifentanil demonstrates potent, naloxone-reversible µ-selective opioid agonist activity in animal assays.206


In animals and humans, remifentanil produces dose-dependent analgesic effects. Human laboratory studies have examined analgesic effects of bolus IV doses (0.0625 to 2.0 µg/kg)207as well as computer-controlled infusions with targeted plasma concentrations (0.75 to 3.0 ng/mL).208 Bolus doses produced a peak analgesic effect between 1 and 3 minutes and a duration of approximately 10 minutes. In volunteers, MEAC is approximately 0.75 ng/mL and analgesic EC50 is approximately 3 ng/mL.208 Both studies found remifentanil to be about 40 times as potent as alfentanil.


Clinical investigations have evaluated early postoperative analgesia. One study reported that after remifentanil–propofol anesthesia, nearly 80% of patients were titrated to satisfactory analgesia with remifentanil infusion of 0.05 to 0.15 µg/kg/min.209 Another early postoperative evaluation demonstrated effective analgesia with patient-controlled infusion of remifentanil to a mean target blood concentration of 2 ng/mL, but noted a fairly high incidence of nausea (26%) with this regimen.210 Clinical evaluations of remifentanil for labor analgesia have produced conflicting results, and some have found prohibitive rates of unacceptable side effects such as nausea and respiratory depression. However, a dose-ranging study that used remifentanil via patient-controlled analgesia reported a median effective bolus dose of 0.4 µg/kg (range, 0.2 to 0.8 µg/kg) and consumption of 0.066 µg/kg/min (range, 0.027 to 0.207 µg/kg/min).211 Although these results are preliminary, remifentanil may offer an alternative for laboring patients in whom regional anesthesia is absolutely contraindicated.

Use in Anesthesia

The effect of remifentanil on the MAC of volatile anesthetics is characterized by steep dose-effect or concentration-effect curves typical of other µ opioid agonists. In animals, remifentanil decreases enflurane and isoflurane MAC in a dose-dependent fashion up to a maximum near 65%, similar to fentanyl.212,213 In humans, remifentanil reduces isoflurane MAC logarithmically in a blood concentration-dependent fashion.214 A whole blood remifentanil concentration of 1.3 ng/mL reduced isoflurane MAC by 50%, with a maximum MAC reduction (91%) at 32 ng/mL. Remifentanil's effects on the MAC-BAR (requirement for blunting the sympathetic response to skin incision) of sevoflurane215 and desflurane216 in 60% N2O are similar. A remifentanil plasma concentration of 1 ng/mL reduced MAC-BAR of the inhalation agents by 60%, while 3 ng/mL decreased MAC-BAR another 30%.

The rapid onset and brief duration of remifentanil suggest that it is suitable for induction of anesthesia. Although a median ED50 of 12 µg/kg for loss of consciousness has been reported, clinical investigations have also found that, as with other opioids, loss of consciousness is not reliably achieved with remifentanil alone, even in doses of 20 µg/kg or more.217,218 Furthermore, a high incidence of muscle rigidity and purposeless movement was seen. Even at 2 µg/kg remifentanil, moderate muscle rigidity was seen in 40% of patients, and at 20 µg/kg, 60% of patients had severe muscle rigidity.217

Drover and Lemmens219 used computer-assisted infusions to determine the blood concentrations of remifentanil required to supplement 66% N2O in patients having abdominal surgery. Other than premedication with 1 to 2 mg midazolam, no sedatives or hypnotics were given. During surgery, the remifentanil EC50 for adequate anesthesia was 4.1 ng/mL for men and 7.5 ng/mL for women. The reason for gender differences in these results was not clear, but could have been related to different types of surgeries. Pediatric patients require twice as much remifentanil as adults (0.15 µg/kg/min vs. 0.08 µg/kg/min) when it is used with propofol for TIVA.220

Investigations of remifentanil for balanced anesthesia, including combination with isoflurane,221,222 sevoflurane,223 and desflurane,224 report similar findings of hemodynamic stability and easy titratability. A clinical trial of remifentanil and desflurane–N2O identified blood remifentanil concentrations that provide an optimal balance between hemodynamic stability and blunting responses to noxious stimulation while permitting rapid recovery.224 In the presence of 2.2 to 2.7% end-tidal desflurane and N2O, optimal remifentanil plasma concentrations were 5 to 7 ng/mL for laryngoscopy and skin closure and 10 ng/mL during abdominal surgery. It is interesting to note that adjustments in remifentanil blunted the sympathetic response to noxious stimulation but did not alter desflurane's effect on the bispectral index analysis of the EEG.

Remifentanil is infused as a component of TIVA more frequently than other opioids. Both remifentanil and propofol can be administered at fixed infusion rates or by computer-controlled systems that provide target plasma concentrations, commonly referred to as target-controlled infusions or TCI. The combination of remifentanil and propofol for TIVA has been used successfully for a variety of inpatient procedures, including coronary artery bypass graft; other major thoracic, neurosurgical, abdominal, and orthopaedic procedures; as well as ambulatory surgery and other painful procedures in adults and children. Two studies demonstrated that a fairly low plasma concentration of remifentanil, TCI at 3.4 to 4 ng/mL, reduces propofol EC50 for intubation by 66%, from approximately 6 to 2 ng/mL,218,225 but further increases in remifentanil dosage only modestly reduced propofol dose requirements, an apparent ceiling effect.225 A clinical dose ranging study found that remifentanil EC50 for laryngoscopy was 14.3 and 1.4 ng/mL with propofol infusions of 44 and 200 µg/kg/min, respectively.226 Response to intubation was prevented in 80% of patients by approximately doubling the remifentanil. A small bolus dose of remifentanil (20 µg) given 30 seconds before induction, can reduce the pain of propofol injection.227

As previously noted, pediatric patients receiving propofol infusion require higher remifentanil doses than adults.

While high-dose remifentanil (1 to 2 µg/kg/min) has been used as a single agent for cardiac anesthesia,228 it is more commonly administered with propofol or isoflurane for “fast-track cardiac anesthesia.” Target remifentanil and propofol concentrations for cardiac surgery228,229 are very similar to those for other procedures with low-dose propofol. In a study comparing remifentanil, sufentanil, and fentanyl for fast-track cardiac anesthesia, Engoren et al.230 found that remifentanil patients were more likely to require treatment for blood pressure fluctuations during and after surgery, but otherwise, the three regimens produced similar outcomes with respect to extubation, intensive care unit stay, and cost.

One drawback of remifentanil use for general anesthesia is that patients require analgesics very soon after an infusion is stopped. A continuation of remifentanil to transition to postoperative analgesia can avoid early pain and accompanying detrimental sympathoadrenal stimulation and is essential for patients undergoing cardiac or other major surgery.

Remifentanil administered by infusion also appears to be useful during monitored anesthesia care for conscious sedation in procedures such as extracorporeal shock wave lithotripsy and colonoscopy,231,232 or in conjunction with regional anesthesia.233,234,235 When compared with propofol, remifentanil provides better analgesia, but results in more nausea and respiratory depression, whereas propofol causes more oversedation. Times required for readiness for discharge are clinically similar. For monitored anesthesia care, the ideal administration regimen appears to be small bolus doses of remifentanil with a continuous infusion combined with low-dose propofol or midazolam.

Other Central Nervous System Effects

Remifentanil produces classic µ opioid agonist effects on the EEG, that is, a concentration-dependent slowing. The plasma concentration associated with 50% maximal EEG changes (EC50) is 15 to 20 ng/mL.236,237 Remifentanil's rapid onset and very short duration results in extremely close tracking of changes in EEG spectral edge with plasma remifentanil concentration.236,237 Like other opioids, remifentanil can produce muscle rigidity, especially with bolus doses. This can be


avoided with using smaller doses and injecting over 60 seconds or more.

Neither remifentanil (0.5 or 1.0 µg/kg) nor alfentanil (10 or 20 µg/kg) given during isoflurane/N2O anesthesia with controlled ventilation affect intracranial pressure, and both produce modest, dose-dependent decreases in mean arterial pressure.238 A multicenter clinical trial comparing remifentanil/N2O with fentanyl/N2O anesthesia found that intracranial pressure and cerebral perfusion pressure were similar with the two regimens.239 In a study comparing cerebrovascular autoregulation in the awake and anesthetized states, remifentanil 0.5µg/kg/min plus propofol preserved cerebral autoregulation, whereas isoflurane 1.8% did not.240

In many cranial and spinal neurosurgical procedures, the ability to monitor motor-evoked potentials (MEPs) is important; opioids, sedative hypnotic drugs, and inhalation agents used in general anesthesia are known to suppress MEPs. A human and animal study compared the effects of phenylpiperidine opioids and hypnotics including thiopental, midazolam, and propofol on MEPs.241 While all opioids and propofol suppressed MEPs in a dose-dependent fashion, remifentanil exerted less suppression than the other opioids and propofol. A target plasma concentration of 9 ng/mL reduced amplitude by 50%, but the quality and reproducibility of MEPs was preserved even at plasma concentration of 15 ng/mL, well within the plasma concentration range that provides surgical anesthesia.

Although remifentanil has not been shown to produce seizure activity, it can be used to reduce methohexital requirement in patients having electroconvulsive therapy. Remifentanil 1 µg/kg allowed a 50% reduction in methohexital dose, which results in seizure prolongation by 50%.242

Respiratory Depression

Remifentanil produces dose-dependent respiratory depression as measured by increases in end-tidal CO2 and decreased oxygen saturation. In a dose-escalation study in normal volunteers, the respiratory depressant effects of remifentanil and alfentanil were compared.207 Peak respiratory depression occurred at 5 minutes after each dose of remifentanil and alfentanil, and the maximal respiratory depressant effect seen after 2 µg/kg remifentanil was similar to that caused by 32 µg/kg alfentanil. The duration of respiratory depression, measured as time to return of blood gases to within 10% of baseline values, was 10 minutes after 1.5 µg/kg and 20 minutes after 2 µg/kg remifentanil compared with 30 minutes after 32 µg/kg alfentanil. During continuous opioid infusion, the ventilatory response to CO2 decreased by approximately 30, 45, and 60% in response to 4-hour remifentanil infusions of 0.025, 0.050, and 0.075 µg/kg/min, respectively.243 Recovery from remifentanil-induced respiratory depression was rapid, and minute ventilation returned to baseline by 8 minutes (range, 5 to 15 minutes) after the infusion was stopped for all infusion rates. In contrast, a 50% decrease in minute ventilation produced by a 4-hour infusion of alfentanil at 0.5 µg/kg/min required 61 minutes (range, 5 to 90 minutes) to return to baseline.243 In a volunteer study, Glass et al.244 reported that the blood remifentanil concentration needed to depress ventilatory response to inspired 8% CO2 by 50% (EC50) was 1.17 ng/mL. Bouillon et al.245 reported a similar EC50 (0.92 ng/mL) and also noted that remifentanil concentrations that are well tolerated at steady state will produce clinically significant respiratory depression when achieved with bolus dosing. In general, clinical comparisons report that respiratory parameters (respiratory rate, O2 saturation, and end-tidal CO2) recover more rapidly after remifentanil compared with other opioids given in equipotent dosage.

Maintenance of spontaneous respiration during general anesthesia with remifentanil and volatile agents or propofol may not be feasible unless low doses of remifentanil are used.246Clinical experience in spontaneously breathing humans receiving remifentanil combined with either isoflurane or propofol demonstrates respiratory depression in 10 to 35% of patients receiving remifentanil at 0.025 µg/kg/min. It increases to nearly 50% in patients receiving 0.05 µg/kg/min and to >90% with remifentanil 0.075 µg/kg/min.247,248 A similar rate of respiratory depression (20%) with need for assisted ventilation is seen in pediatric patients receiving remifentanil/propofol infusions for general anesthesia during bone marrow aspiration.249 As discussed earlier, remifentanil alone or combined with low-dose propofol or midazolam can be used for conscious sedation and to supplement regional or local anesthesia during monitored anesthesia care. Clinical reports describing these regimens report respiratory depression (respiratory rate <8 or SpO2 <90%) in 2 to 30% of patients, but in all cases, recovery from respiratory depression with remifentanil is more rapid than other agents.231,232,233,234 As with other opioids, higher rates of respiratory depression are seen when propofol is combined with remifentanil (15 to 50% of patients), and careful monitoring and titration are required to minimize this side effect.

Hemodynamic Effects

In healthy volunteers, remifentanil in bolus doses >1.0 µg/kg produce brief increases in systolic blood pressure (5 to 20 torr) and heart rate (10 to 25 beats/min).207 In patients anesthetized with isoflurane and 66% N