1246

Opioids

Opioid analgesics are extensively used for premedication, as a supplement to regional and general anesthesia, as the primary anesthetic agent, and as an analgesic for postoperative pain (see Chapter 72 ). The analgesia produced by these drugs through specific receptor systems within the CNS decreases autonomic, endocrine, and somatic responses to noxious stimulation. Although opioids have been used as sole anesthetics, they create incomplete hypnotic effects at very large doses. Opioids need to be combined with hypnotic drugs to induce the anesthetic state.

Opioids as Complete Anesthetics

In 1947, Neff and colleagues^[103] used meperidine as an intravenous supplement to nitrous oxide/oxygen anesthesia in what is now known as a "balanced anesthesia" technique. The later use of opioids as complete anesthetics coincided with the development of cardiac surgery and intensive care during the 1970s. Providing anesthesia for patients with severe valvular or congenital heart disease without causing cardiovascular collapse was problematic before the use of opioids. These early cardiac surgery patients were extremely ill and had little or no circulatory reserve. During the late 1960s, Lowenstein and associates^[104] noted the hemodynamic stability of patients undergoing mechanical ventilation who were given frequently large doses of intravenous morphine to suppress respiration in intensive care units. This observation encouraged Lowenstein and coworkers to become the first to administer morphine (0.5 to 3.0 mg/kg) as the only anesthetic drug. The resulting cardiovascular stability in acutely ill patients with acquired valvular heart disease was impressive. As cardiac surgery advanced in methodology, patients with ischemic heart disease began to undergo surgical anesthesia. Unfortunately, morphine anesthesia was less satisfactory for these patients, in whom hypertension, tachycardia, and awareness during surgery developed, in contrast to patients with valvular heart disease.^[104]

In 1978, Stanley and Webster^[105] introduced the concept of high-dose fentanyl for cardiac anesthesia. This technique minimized the undesirable effects of morphine on induction (hypotension) and provided better hemodynamic stability in patients with good ventricular function and ischemic heart disease. As clinical experience with fentanyl increased, however, investigators found that even increasingly large doses of fentanyl could not always produce complete unresponsiveness, as defined in Figure 31-8 , for the most potent stimuli and difficult-to-suppress responses. ^[106] This discovery raised the important issue of whether opioids are "complete anesthetics" or are, as suggested in Figure 31-3 , only potent analgesics and weak hypnotics.

Human and animal studies show that opioids are not complete anesthetics. Wynands and colleagues^[21] used moderate to large doses of fentanyl (50 to 150 µg/kg) only, with no other anesthetic drugs (no hypnotic component), and measured plasma concentrations at defined surgical stimuli (intubation, skin incision, sternotomy, aortic root dissection) in patients with good ventricular function who were undergoing coronary surgery. Figure 31-16 shows the relationship among drug concentrations, stimulation,

Figure 31-16 Plasma concentration of fentanyl at the time of certain clinical stimuli, along with the presence or absence of hypertension. During sternotomy and aortic dissection, even high plasma concentrations of fentanyl (>15 ng/mL) do not always prevent hypertension. (Redrawn with modification from Wynands JE, Wong P, Townsend GE, et al: Narcotic requirements for intravenous anesthesia. Anesth Analg 63:101, 1984.)

The investigations conducted by Murphy and Hug^[110] in measuring the depth of opioid anesthesia in animals (dogs) also address the issue of whether opioids are complete anesthetics. These investigators examined the ability of fentanyl to decrease enflurane MAC. They first anesthetized the dog with enflurane and determined MAC. Several infusions of fentanyl at progressively higher rates were used to obtain a constant steady-state plasma concentration of fentanyl in each animal. After each increase in infusion rate, enflurane MAC was determined again. Measurement of opioid concentrations in blood samples ensured that several different steady-state plasma concentrations of fentanyl were obtained in each animal. Murphy and Hug found that even high plasma concentrations of fentanyl (20 ng/mL) did not decrease enflurane MAC beyond 60% to 70% of its initial value ( Fig. 31-17 ). That is, there was a ceiling to the enflurane-sparing effect. Morphine, sufentanil, and alfentanil also decrease enflurane MAC and have a similar ceiling effect in dogs.^{[111] [112] [113]}

McEwan and colleagues undertook similar studies in humans and characterized the decrease in isoflurane MAC with varying constant fentanyl plasma concentrations achieved with a computer-driven infusion pump.^[23] Movement response to initial skin incision was examined

1247

Figure 31-17 Percent reduction in enflurane minimum alveolar concentration (MAC) as a function of plasma concentrations of fentanyl. Each point represents the mean concentration (±SEM) for fentanyl in plasma and the average percent (±SEM) reduction in enflurane MAC. Numbers of dogs below the vertical standard error bars indicate the numbers per data point. (Redrawn from Murphy MR, Hug CC Jr: The anesthetic potency of fentanyl in terms of its reduction of enflurane MAC. Anesthesiology 57:485, 1982.)

Comparable results are found when other end points of inhaled anesthetics and opioid interactions are examined, including MAC_awake and MAC_BAR ;^{[36] [37] [116] [117]} a ceiling effect of the opioid occurs at defined steady-state plasma concentrations.

Clinical Signs of Inadequate Anesthesia and Plasma Concentration of Opioids

Ausems and colleagues^[118] used pharmacodynamic modeling concepts to relate clinical signs of inadequate opioid anesthesia to plasma concentrations of the drug. In their study paradigm, patients were premedicated with a benzodiazepine, anesthesia was induced with alfentanil (150 µg/kg), and the trachea was intubated with the aid of succinylcholine. Anesthesia was maintained with 70% nitrous oxide and a variable-rate infusion of alfentanil. The infusion was titrated to the following clinical end points: (1) increased systemic arterial blood pressure greater than 15 mm Hg higher than the patient's normal value; (2) heart rate exceeding 90 beats/min in the absence of hypovolemia; (3) somatic responses such as body movements (minimal muscle paralysis allowed physical movement), swallowing, coughing, grimacing, or opening of the eyes; and (4) autonomic signs of inadequate anesthesia (e.g., lacrimation, flushing, or sweating). If any clinical signs occurred, the infusion rate was increased 25 to 50 µg/kg/hr, and a small bolus dose (7 µg/kg) was given. Good hemodynamic control was possible in all subjects. If no clinical signs occurred, however, the infusion rate was decreased at regular 15-minute intervals.

Table 31-5 shows the incidence of response to intraoperative noxious stimuli in 37 patients; three different types of surgical procedures are represented. Although hypertension was the most common clinical response, the other three clinical measures occurred a significant number of times. The authors could not predict which patients would have somatic responses, tachycardia, or hypertension. Measurement of plasma concentrations made it possible to describe the concentration-versus-response relationship for different perioperative stimuli. Figure 31-18A shows the relationship between the plasma concentration of alfentanil and response/no response for three clinical end points: intubation, skin incision, and skin closure. Regarding elimination of the response to noxious stimulation, intubation required significantly higher plasma concentrations of alfentanil than skin incision did, and skin closure required significantly lower concentrations than skin incision did.

1248

Figure 31-18 A, Relationship between the plasma concentration of alfentanil and response/no response at three specific events of short duration. The quantal data are characterized by logistic regression in the lower panel, ±SE for the Cp₅₀ , which is the alfentanil plasma concentration associated with a 50% probability of no response. (From Ausems ME, Hug CC Jr, Stanski DR, et al: Plasma concentrations of alfentanil required to supplement nitrous oxide anesthesia for general surgery. Anesthesiology 65:362, 1986.) B, Plasma concentration of alfentanil versus the probability of no response for each of 34 patients during the intra-abdominal phase of lower abdominal surgery. Dots represent the Cp₅₀ values, and the heavy dark line represents the average response of the 34 patients. (From Ausems ME, Vuyk J, Hug CC Jr, et al: Comparison of computer-assisted infusion versus intermittent bolus administration of alfentanil as a supplement to nitrous oxide for lower abdominal surgery. Anesthesiology 68:851, 1988.)

The use of logistic regression made it possible to define Cp₅₀ values for these clinical events ( Table 31-6 ). Plasma concentrations of alfentanil were varied in the individual subject during surgery to obtain multiple response/no response data points, after which a curve plotting plasma concentration against response was constructed for each

TABLE 31-6 -- Cp₅₀ values for perioperative events and intraoperative manipulation associated with three types of surgical procedures during alfentanil anesthesia

Event	Cp50 (ng/mL)
Single events
Intubation	475 ± 28 *
Skin incision	279 ± 20
Skin closure	150 ± 23
Spontaneous ventilation	233 ± 13
Intraoperative manipulation
Breast surgery	270 ± 63 †
Lower abdominal	309 ± 44
Upper abdominal	412 ± 35
Spontaneous ventilation	233 ± 13
Cp50 , plasma concentration of a drug producing a 50% chance of suppressing a response to a given stimulus.
Data from Ausems ME, Vuyk J, Hug CC Jr, et al: Comparison of computer-assisted infusion versus intermittent bolus administration of alfentanil as a supplement to nitrous oxide for lower abdominal surgery. Anesthesiology 68:851, 1988.

*Standard error of the estimated value for the parameter.
†Values are presented as mean ± SD.

The foregoing pharmacodynamic modeling concept has been used to examine the alfentanil dose requirement, pharmacokinetics, and pharmacodynamics in patients with heavy alcohol consumption. Lemmens and colleagues^[119] showed that moderate consumers of alcohol (20 to 40 g/day) have an increased requirement for alfentanil because of decreased CNS sensitivity to this opioid. The Cp₅₀ of alfentanil in patients with alcohol consumption was 522 ± 104 ng/mL versus 208 ± 26 ng/mL in a control group. This finding suggests that there may be cross-tolerance between opioids and ethanol. Lemmens and coworkers^[120] also defined the role of plasma protein binding in determining the clinical effects of alfentanil anesthesia. By examining the alfentanil plasma concentration-versus-clinical response relationship, both total bound and free, and the degree of alfentanil plasma protein binding, they found a significant negative correlation (r = .67) between Cp₅₀ (total drug) and the free fraction of alfentanil. These data suggest that 45% of the variability in alfentanil CNS response, as estimated by the Cp₅₀ , could be explained by variability in the plasma protein binding of alfentanil.

Remifentanil is a newer opioid that is related to the fentanyl family of short-acting, 4-aniliopiperidine derivatives; however, it is unique in having ester linkages that are susceptible to hydrolysis by nonspecific blood and tissue esterases, thereby resulting in very rapid metabolism.

1249

The rapid blood-brain equilibration of alfentanil and remifentanil means that a given plasma concentration has a close relationship to the CNS concentration and hence to drug effect.^[125] Drugs with a slower rate of blood-brain equilibration (e.g., fentanyl, sufentanil, morphine) would be less amenable to the type of pharmacodynamic analysis of the plasma concentration-versus-clinical effect relationship used by Ausems and coworkers.^[39] Glass and colleagues^[126] applied the fundamental concepts proposed by Ausems and coworkers^[39] to fentanyl by using a computer-driven infusion pump to obtain constant fentanyl plasma concentrations from induction of anesthesia to skin incision. Because of the long equilibration delay between plasma and the effect site in the CNS (t_½ k_e0 of 4.7 minutes), a significant period of time was allowed between establishing the constant plasma concentration and application of the noxious stimuli. Only fentanyl and 70% nitrous oxide were used, no premedication or induction sedative was given, and fentanyl plasma concentrations were also measured. After skin incision, patients were observed for purposeful somatic movement. The fentanyl plasma concentration that had a 50% probability of no movement (Cp₅₀ ) with 70% nitrous oxide was 3.26 ng/mL.

The approach to opioid administration presented by Ausems and colleagues^[39] provides useful insight into clinical assessment of the analgesic component of depth of anesthesia. Overdosage with opioids cannot be judged intraoperatively. Only at the end of anesthesia, when spontaneous ventilation should occur, does one know whether administration of opioids has been excessive. To prevent overdosage, Ausems and coworkers^[39] proposed using a viable rate of infusion in which one titrates plasma concentrations to clinical effect to find the lowest possible effective rate of opioid administration. To achieve this end point, one must look for clinical signs of inadequate anesthesia. Once the infusion rate just causing inadequate anesthesia has been determined, one increases the rate slightly, thus providing adequate anesthesia. The steep slope of the concentration-versus-response curve for alfentanil (see Fig. 31-18B ) demonstrates that a small increase in plasma concentration of the drug rapidly converts inadequate anesthesia (100% probability of response to stimuli) to adequate anesthesia. Because of moderate variability in the pharmacokinetics of alfentanil, this process of titrating the dose (infusion rate) against clinical effect and pharmacodynamics is necessary for each patient. This concept is most applicable to alfentanil, which has rapid blood-brain equilibration,^[125] and it is ideal with remifentanil, which has extremely rapid and predictable pharmacokinetics along with rapid blood-brain equilibration.^[121] Intermittent intravenous bolus administration of opioids is not as efficient as variable-rate infusion when titrating plasma concentrations of opioids to clinical effect.^[118]