Physicochemical Properties

Sodium citrate is an alkali and neutralizes acid; it is often administered orally to reduce the likelihood of pneumonitis after regurgitation of gastric contents. Chelating agents (chel is the Greek word for a crab’s claw) combine chemically with metal ions, reducing their toxicity and enhancing elimination, usually in the urine. Such drugs include desferrioxamine (chelates iron and aluminium), dicobalt edetate (cyanide toxicity), sodium calcium edetate (lead) and penicillamine (copper and lead). Stored blood contains a citrate-based anticoagulant which prevents clotting; this chelates calcium ions and may cause hypocalcaemia after massive blood transfusion. Phenol and alcohol denature proteins; they are used occasionally to produce prolonged or permanent nerve blockade.

Action on Receptors

A receptor is a complex structure on the cell membrane which can bind selectively with endogenous compounds or drugs, resulting in changes within the cell which modify its function. These include changes in selective ion channel permeability (e.g. acetylcholine, glutamate, GABA receptors), cyclic adenosine monophosphate (e.g. opioid, β, α₂ and dopamine receptors), cyclic guanosine monophosphate (e.g. atrial natriuretic peptide receptor), inositol phosphate and diacylglycerol (e.g. α₁, angiotensin AT₁, endothelin, histamine H₁ and vasopressin V₁ receptors) and nitric oxide (e.g. muscarinic M₃ receptor).

A compound which binds to a receptor and changes intracellular function is termed an agonist. The classic dose–response relationship of an agonist is shown in Figure 1.1. As the concentration of the agonist increases, a maximum effect is reached as the receptors in the system become saturated (Fig. 1.1A). Conventionally, log dose is plotted against effect, resulting in a sigmoid curve which is approximately linear between 20 and 80% of maximum effect (Fig. 1.1B). Three agonists are shown in Figure 1.2. Agonist A produces 100% effect at a lower concentration than agonist B. Therefore, compared with A, agonist B is less potent but has similar efficacy. Drug C is termed a partial agonist as the maximum effect is less than that of A or B. Buprenorphine is a partial agonist (at the μ-opioid receptor), as are some of the β-blockers with intrinsic activity, e.g. oxprenolol, pindolol, acebutalol, celiprolol.

Antagonists combine selectively with the receptor but produce no effect. They may interact with the receptor in a competitive (reversible) or non-competitive (irreversible) fashion. In the presence of a competitive antagonist, the dose–response curve of an agonist is shifted to the right but the maximum effect remains unaltered (Fig. 1.3A). Examples of this effect include the displacement of morphine by naloxone and endogenous catecholamines by β-blockers.

A non-competitive (irreversible) antagonist also shifts the dose–response curve to the right but, with increasing concentrations, reduces the maximum effect (Fig. 1.3B). For example, the α₁-antagonist phenoxybenzamine, used in the preoperative preparation of patients with phaeochromocytoma, has a long duration of action because of the formation of stable chemical bonds between drug and receptor.

The relationship between drug dose and response is often described by a Hill plot (Fig. 1.4). A typical agonist such as that shown in Figure 1.1 produces a straight line with a slope (i.e. Hill coefficient) of + 1.

Action on Enzymes

Drugs may act by inhibiting the action of an enzyme or competing for its endogenous substrate. Reversible inhibition is the mechanism of action of edrophonium (acetylcholinesterase), aminophylline (phosphodiesterase) and captopril (angiotensin-converting enzyme). Irreversible enzyme inhibition occurs when a stable chemical bond is formed between drug and enzyme, resulting in prolonged or permanent inactivity e.g. omeprazole (gastric hydrogen-potassium ATPase), aspirin (cyclo-oxygenase) and organophosphorus compounds (acetylcholinesterase).

However, the interaction between drug and enzyme may be more complex than this simple classification implies. For example, neostigmine inhibits acetylcholinesterase in a reversible manner, but the mechanism of action is more akin to that of an irreversible drug because neostigmine forms covalent chemical bonds with the enzyme.

Enzyme Induction and Inhibition

Some drugs may enhance the activity of enzymes responsible for drug metabolism, particularly the cytochrome P450 enzymes and glucuronyl transferase. Such drugs include phenytoin, carbamazepine, phenylbutazone, barbiturates, ethanol, steroids and some inhalational anaesthetic agents (halothane, enflurane). Cigarette smoking also induces cytochrome P450 enzymes.

Drugs with mechanisms of action other than on enzymes may also interfere significantly with enzyme systems. For example, etomidate inhibits the synthesis of cortisol and aldosterone – an effect which may explain the increased mortality in critically ill patients which occurred when it was used as a sedative in intensive care. Cimetidine is a potent enzyme inhibitor and may prolong the elimination of drugs such as diazepam, propranolol, oral anticoagulants, phenytoin and lidocaine. Troublesome interactions with enzyme systems are less of a problem with new drugs; if significant enzyme interaction is discovered in the early stages of development, the drug is usually abandoned.

Volume of Distribution

Volume of distribution is a good example of the abstract nature of pharmacokinetics; it is not a real volume but merely a concept which helps us to understand what we observe. Nevertheless, it is a very useful notion which enables us to predict certain properties of a drug and also calculate other pharmacokinetic values.

Imagine that a patient receiving an intravenous dose of an anaesthetic induction agent is a bucket of water and that the drug is distributed evenly throughout the water immediately after injection. The volume of water represents the initial volume of distribution (V). It may be calculated easily:

(1)

where C₀ is the initial concentration. Therefore:

(2)

A more accurate measurement of V is possible during constant rate infusion when the distribution of the drug in the tissues has time to equilibrate; this is termed volume of distribution at steady state (V_ss).

Drugs which remain in the plasma and do not pass easily to other tissues have a small V and therefore a large C₀. Relatively ionized drugs (e.g. muscle relaxants) or drugs highly bound to plasma proteins (e.g. NSAIDs) often have a small V. Drugs with a large V are often lipid-soluble and therefore penetrate and accumulate in tissues outside the plasma (e.g. intravenous induction agents). Some drugs accumulate outside the plasma, making values for V greater than total body volume (a reminder of the abstract nature of pharmacokinetics). Large V values are often observed for drugs highly bound to proteins outside plasma (e.g. local anaesthetics, digoxin).

Several factors may affect V and therefore C₀ on bolus injection of a drug. Patients who are dehydrated, or have lost blood, have a significantly greater plasma C₀ after a normal dose of intravenous induction agent, increasing the likelihood of severe side-effects, especially hypotension. Neonates have a proportionally greater volume of extracellular fluid compared with adults, and water-soluble drugs (e.g. muscle relaxants) tend to have a proportionally greater V. Factors affecting plasma protein binding (see above) may also affect V.

Finally, V can give some indication as to the half-life. A large V is often associated with a relatively slow decline in plasma concentration; this relationship is expressed below in a useful pharmacokinetic equation (eqn 4).

Clearance

Clearance is defined as the volume of blood or plasma from which the drug is removed completely in unit time. Drugs may be eliminated from the blood by the liver, kidney or occasionally other routes (see above). The relative proportion of hepatic and renal clearance of a drug is important. Most drugs used in anaesthetic practice are cleared predominantly by the liver, but some rely on renal or non-organ-dependent clearance. Excessive accumulation of a drug occurs in patients in renal failure if its renal clearance is significant. For example, morphine is metabolized primarily in the liver and this is not affected significantly in renal impairment. However, the active metabolite morphine-6-glucuronide is excreted predominantly by the kidney. This accumulates in renal insufficiency and is responsible for increased morphine sensitivity in these patients.

As with volume of distribution, clearance may suggest likely properties of a drug. For example, if clearance is greater than hepatic blood flow, factors other than hepatic metabolism must account for its total clearance. Values greater than cardiac output may indicate metabolism in the plasma (e.g. succinylcholine) or other tissues (e.g. remifentanil). Clearance is an important (but not the only) factor affecting t_1/2 and steady-state plasma concentrations achieved during constant rate infusions (see below).

Elimination Half-Life

Methods of administration of a drug are influenced considerably by its plasma t_1/2, as this often reflects duration of action. It is important to remember that t_1/2 is influenced not only by clearance (Cl) but also by V:

(3)

or

The constant in this equation (elimination rate constant) is the natural logarithm of 2 (ln 2) i.e. 0.693. Therefore:

(4)

Half-life often reflects duration of action but not if the drug acts irreversibly (e.g. some NSAIDs, omeprazole, phenoxybenzamine) or if active metabolites are formed (Table 1.1).

So far, we have considered metabolic or elimination t_1/2 only. The initial decrease in plasma concentrations after administration of many drugs, especially if given intravenously, occurs primarily because of redistribution into tissues. Therefore, the simple relationship between elimination t_1/2 and duration of action does not apply in many situations (see below, ‘Two-Compartment Models’).

Calculating t_1/2, V and Clearance

It is a simple exercise to calculate these values for a drug after intravenous bolus administration. A known dose is given and regular blood samples are taken for plasma concentration measurements. In this example, we assume that the drug remains in the plasma and is removed only by metabolism; this is a called a one-compartment model. After achieving C₀, plasma concentration (C_P) declines in a simple exponential manner as shown in Figure 1.5A. If the natural logs of the concentrations are plotted against time (semilog plot), a straight line is produced (Figure 1.5B). The gradient of this line is the elimination rate constant k, which is related to t_1/2 in the following equation:

(5)

We may calculate V using equation (2) and then clearance from equation (4). C_P may be predicted at any time from the following equation:

(6)

where t is the time after administration.

Clearance may be derived also by calculation of the area under the concentration-time curve extrapolated to infinity (AUC_∞) and substitution in the following equation:

(7)

Two-Compartment Models

The body is not, of course, a single homogeneous compartment; drug plasma concentrations are the result of elimination by metabolism and redistribution to and from tissues such as brain, heart, liver, muscles and fat. The mathematics describing this real situation are extremely complex. However, plasma concentrations of many drugs behave approximately as if they were distributed in two or three compartments. Applying these mathematical models is a reasonable compromise.

Let us consider a two-compartment model; one compartment may be thought of as representing the plasma and the other, the remainder of the body. When an intravenous bolus is injected into this system, C_P decreases because of an exponential decay resulting from elimination and another exponential decay resulting from redistribution into the tissues. Therefore, when C_P is plotted against time, the curve may be described by a biexponential equation. If plotted on a semilogarithmic plot (Fig. 1.6), two straight lines can be identified and derived. Their gradients are the elimination rate constants dependent on elimination (β) and redistribution (α).

Redistribution kinetics are not only of theoretical interest, because it is often the decline in C_P resulting from redistribution which is responsible for the cessation of an observed effect of a drug; intravenous induction agents and initial doses of intravenous fentanyl are good examples of this. Patients wake up after a bolus administration of propofol because of redistribution, not metabolism.

Calculating the separate pharmacokinetic values is easy; one curve is simply subtracted from the other. Consider Figure 1.6 in which natural log concentration is plotted against time and two slopes are seen. The second and less steep slope represents decline in plasma concentration caused by elimination of the drug by metabolism. From this, the elimination half-life (t_1/2^β) may be calculated. In order to calculate the half-life of the redistribution phase (t_1/2^α), the elimination slope is extrapolated back to time 0. If data on this imaginary part of the elimination slope are subtracted from those on the real line above it, another imaginary line may be constructed which represents that part of the decline in plasma concentration which is the result of redistribution. From this line, the redistribution half-life (t_1/2^α) may be calculated.

The equation for C_P at any time in a two-compartment model after bolus intravenous administration is therefore:

(8)

where α and β are the redistribution and elimination rate constants, respectively, and A and B are values derived by back extrapolation of the redistribution and elimination slopes to the y-axis.

Some drugs, e.g. propofol, are best fitted to a triexponential, three-compartment model which reveals half-lives for two processes of redistribution (conventionally t_1/2^α and t_1/2^β) and one for elimination (t_1/2^γ). Equations developed from these basic concepts are contained in the software of target-controlled infusion pumps.

Context-Sensitive Half-Life

This concept refers to plasma half-life (time for plasma concentration to decline by 50%) after an intravenous drug infusion is stopped; ‘context’ refers to the duration of infusion. The amount of drug accumulating in body tissues increases with duration of infusion for most drugs. Consequently, on stopping the infusion, time for the plasma concentration to decline by 50% depends on duration of infusion. The longer the infusion, the more drug accumulates and the longer the plasma half-life becomes, because there is more drug to enter the plasma on stopping the infusion.

Figure 1.7 shows the effect of infusion duration on the half-lives of alfentanil, fentanyl and remifentanil. Alfentanil, and especially fentanyl, accumulate during infusion, causing an increase in context-sensitive half-life as the duration of infusion increases. In other words, time to recovery from alfentanil- or fentanyl-based anaesthesia depends on duration of infusion. Remifentanil is metabolized by tissue esterases and does not accumulate. Therefore, time for plasma concentration of remifentanil to decline by 50% is independent of duration of infusion, i.e. recovery times after remifentanil-based anaesthesia are short and predictable, no matter how long the infusion has run.

Oral

The oral route of drug administration is important in modern anaesthetic practice (e.g. premedication, postoperative analgesia). It is often necessary also to continue concurrent medication during the perioperative period (e.g. antihypertension therapy, anti-anginal medication). It is therefore important to appreciate the factors involved in the absorption of orally administered drugs.

The formulation of tablets or capsules is very precise, as their consistent dissolution is necessary before absorption can take place. The rate of absorption, and therefore effect of the drug, may be influenced significantly by this factor. Most preparations dissolve in the acidic gastric juices and the intact drug is absorbed in the upper intestine. However, some drugs are broken down by acids (e.g. omeprazole, benzylpenicillin) or are irritant to the stomach (e.g. aspirin, phenylbutazone) and may be given as enteric-coated preparations. Drugs given in solution are often absorbed more rapidly but this may induce nausea or vomiting immediately after anaesthesia. Some drugs used in anaesthetic practice are available in slow-release preparations (e.g. morphine, oxycontin, tramadol).

Lingual and Buccal

This is a useful method of administration if a drug is lipid-soluble and crosses the oral mucosa with relative ease. First-pass metabolism is avoided. Glyceryl trinitrate and buprenorphine are available as sublingual tablets and morphine as a buccal preparation.

Intramuscular

Intramuscular administration is still used occasionally in the perioperative period. It may avoid the problems associated with large initial plasma concentrations after rapid intravenous administration, is devoid of first-pass effects and may be administered relatively easily. However, absorption may be unpredictable, some preparations are particularly painful and irritant (e.g. diclofenac) and complications include damage to nervous and vascular tissue and inadvertent intravenous injection. It is disliked intensely by most adults and nearly all children.

Variations in absorption may be clinically relevant. For example, peak plasma concentrations of morphine may occur at any time from 5 to 60 min after intramuscular administration, an important factor in the failure of this method to produce good reliable analgesia (see Ch 41).

Subcutaneous

Absorption is very susceptible to changes in skin perfusion, and tissue irritation may be a significant problem. However, this method is used in many centres for providing postoperative pain relief, particularly in children or after intermediate surgery, and has the advantage that potentially difficult intravenous access is not required. A small cannula is placed subcutaneously during anaesthesia and can be replaced, if necessary, with relative ease. Even patient-controlled analgesia (PCA) has been used effectively by this route.

Intravenous

Rectal

This technique reduces the problems of first-pass metabolism and the need for injections. It is used in children and adults (paracetamol, diclofenac, ibuprofen) for postoperative analgesia. However, the proportion of drug absorbed is very variable.

Transdermal

Drugs with a high lipid solubility and potency may be given transdermally. The pharmacological properties of glyceryl trinitrate render it ideal for this technique (i.e. potent, highly lipid-soluble, short half-life). Transdermal hyoscine is used for travel and other causes of sickness. Fentanyl patches can be very effective, particularly in patients with cancer pain. Buprenorphine and lidocaine transdermal delivery systems are also available. The latter is used for post-herpetic neuralgia which has not responded to more conventional techniques.

It may take some time before a steady-state plasma concentration is achieved and many devices incorporate large amounts of drug in the adhesive layer in order to provide a loading dose which reduces this period. At steady state, transdermal delivery has several similarities to intravenous infusion. However, on removing the adhesive patch, plasma concentrations may decline relatively slowly because of a depot of drug in the surrounding skin; this occurs with transdermal fentanyl systems.

Inhalation

The delivery of inhaled volatile anaesthetics is discussed below, but other drugs may be given by this route, especially bronchodilators and steroids. Atropine and Adrenaline are absorbed if injected into the bronchial tree and this offers a route of administration in emergencies if no other method of delivery is possible. Opioids such as fentanyl and diamorphine have been given as nebulized solutions but this technique is not routine.

Epidural

This is a common route of administration in anaesthetic practice. The epidural space is very vascular and significant amounts of drug may be absorbed systemically, even if any vessels are avoided by the needle or cannula. Opioids diffuse across the dura to act on spinal opioid receptors, but much of their action when given epidurally is the result of systemic absorption. Complications include haematoma and infection, inadvertent dural puncture with consequent headache or spinal administration of the drug.

Spinal (Subarachnoid)

When given spinally, drugs have free access to the neural tissue of the spinal cord and small doses have profound, rapid effects, an advantage and also a disadvantage of the method. Protein binding is not a significant factor because CSF protein concentration is relatively low.

Pharmaceutical

In this type of interaction, drugs mixed in the same syringe or infusion bag react chemically with adverse results. For example, mixing succinylcholine with thiopental (pH 10–11) hydrolyses the former, rendering it inactive. Before mixing drugs, data should be sought on their compatibility.

Pharmacokinetic

Absorption of a drug, particularly if given orally, may be affected by other drugs because of their action on gastric emptying (see above). Interference with protein binding (see above) is a common cause of drug interaction. We have discussed drug metabolism in some detail and there are many potential sites in this process where interactions can occur (e.g. competition for enzyme systems, enzyme inhibition or induction).

Pharmacodynamic

This is the most frequent type of interaction in anaesthetic practice. A typical anaesthetic is a series of pharmacodynamic interactions. These may be adverse (e.g. increased respiratory depression with opioids and volatile agents) or advantageous (e.g. reversal of muscle relaxation with neostigmine). An understanding of the many subtle pharmacodynamic interactions in modern anaesthesia accounts for much of the difference in the quality of anaesthesia and recovery associated with the experienced compared with the novice anaesthetist.

Mechanism of Action

The exact mechanism of action of volatile anaesthetic agents is at present unknown. Potency is, in general, related to lipid solubility (Meyer-Overton relationship, Table 1.3) and this has given rise to the concept of volatile agents dissolving in the lipid cell membrane in a non-specific manner, disrupting membrane function and thereby influencing the function of proteins, e.g. ion channels. However, it is now appreciated that volatile agents affect neuronal function as a consequence of binding to specific protein sites (e.g. GABA_A receptor).

Potency

The potency of volatile agents is defined in terms of minimum alveolar concentration (MAC). MAC is the alveolar concentration of a volatile agent which produces no movement in 50% of spontaneously breathing patients after skin incision. MAC is inversely related to lipid solubility (Table 1.3).

Onset of Action

When considering onset of action of volatile agents, there is a fundamental difference compared with intravenous agents. Effects of non-volatile drugs are related to plasma or tissue concentrations; this is not so with volatile agents. Partial pressure of the volatile agent is important, not concentration. If a volatile agent is highly soluble in blood, partial pressure increases slowly because large amounts dissolve in the blood. Consequently, onset of anaesthesia is slow with agents soluble in blood and rapid with agents which are relatively insoluble. The same applies to recovery from anaesthesia. Table 1.4 lists the most commonly used inhaled agents (in order of speed of onset) and their relative blood/gas solubilities.

Alveolar partial pressure (P_A) is assumed to be equivalent to cerebral artery partial pressure and therefore depth of anaesthesia. At a fixed inspired partial pressure (P_I), the rate at which P_A approaches P_I is related to speed of onset of effect (Fig. 1.9). This is rapid with agents of low blood solubility (e.g. sevoflurane) and relatively slow with more soluble agents (e.g. halothane).

Clearly, solubility of the agent in blood is a major determinant of the speed of onset of anaesthesia, but other factors can have significant effects. The rate of delivery of the agent to the alveoli is important; therefore increasing P_I by adjusting the vaporizer (a factor limited with some agents by irritant effects on the airway in spontaneously breathing patients), reducing apparatus dead space and increasing alveolar ventilation increase speed of induction of anaesthesia. If cardiac output is reduced, relatively less agent is removed from the alveolus and P_A increases towards P_I more rapidly. Consequently, induction of anaesthesia is more rapid in patients with reduced cardiac output. Both rate of delivery and cardiac output have particularly significant effects with agents that are relatively soluble in blood but less so with insoluble agents.

Ventilation/perfusion mismatch may reduce the speed of induction, an effect more significant in agents of low solubility. For example, if one lung is collapsed (i.e. perfused but not ventilated) increasing ventilation or inspired concentration of agents such as halothane helps to compensate. However, this is not the case for agents such as sevoflurane.

Calvey, N., Williams, N. Principles and practice of pharmacology for anaesthetists, fifth ed. Oxford: Wiley-Blackwell; 2008.