Irreversible inhibition occurs with organophosphorus insecticides and chemical warfare agents (see Chap. 10), which combine covalently with the active site of acetylcholinesterase; recovery of cholinesterase activity depends on the formation of new enzyme. Covalent binding of aspirin to cyclo-oxygenase (COX) inhibits the enzyme in platelets for their entire lifespan because platelets have no system for synthesising new protein; this is why low doses of aspirin are sufficient for antiplatelet action.

Selectivity

The pharmacologist who produces a new drug and the doctor who gives it to a patient share the desire that it should possess a selective action so that additional and unwanted (adverse) effects do not complicate the management of the patient. Approaches to obtaining selectivity of drug action include the following.

Modification of drug structure

Many drugs have in their design a structural similarity to some natural constituent of the body, e.g. a neurotransmitter, a hormone, a substrate for an enzyme; replacing or competing with that natural constituent achieves selectivity of action. Enormous scientific effort and expertise go into the synthesis and testing of analogues of natural substances in order to create drugs capable of obtaining a specified effect, and that alone (see Therapeutic index, below). The approach is the basis of modern drug design and it has led to the production of adrenoceptor antagonists, histamine receptor antagonists and many other important medicines.

But there are biological constraints to selectivity. Anticancer drugs that act against rapidly dividing cells lack selectivity because they also damage other tissues with a high cell replication rate, such as bone marrow and gut epithelium.

Selective delivery (drug targeting)

Simple topical application, e.g. skin and eye, and special drug delivery systems, e.g. intrabronchial administration of β₂-adrenoceptor agonist or corticosteroid (inhaled, pressurised, metered aerosol for asthma) can achieve the objective of target tissue selectivity. Selective targeting of drugs to less accessible sites of disease offers considerable scope for therapy as technology develops, e.g. attaching drugs to antibodies selective for cancer cells.

Stereoselectivity

Drug molecules are three-dimensional and many drugs contain one or more asymmetrical or chiral¹ centres in their structures, i.e. a single drug can be, in effect, a mixture of two non-identical mirror images (like a mixture of left- and right-handed gloves). The two forms, which are known as enantiomorphs, can exhibit very different pharmacodynamic, pharmacokinetic and toxicological properties.

For example, (1) the S form of warfarin is four times more active than the R form,² (2) the peak plasma concentration of S fenoprofen is four times that of R fenoprofen after oral administration of RS fenoprofen, and (3) the S, but not the R, enantiomorph of thalidomide is metabolised to primary toxins.

Many other drugs are available as mixtures of enantiomorphs (racemates). Pharmaceutical development of drugs as single enantiomers rather than as racemic mixtures offers the prospect of greater selectivity of action and lessens risk of toxicity.

Quantitative aspects

That a drug has a desired qualitative action is obviously all important, but is not by itself enough. There are also quantitative aspects, i.e. the right amount of action is required, and with some drugs the dose has to be adjusted very precisely to deliver this, neither too little nor too much, to escape both inefficacy and toxicity, e.g. digoxin, lithium, gentamicin. While the general correlation between dose and response may evoke no surprise, certain characteristics of the relation are fundamental to the way drugs are used, as described below.

Dose–response relationships

Conventionally, the horizontal axis shows the dose and the response appears on the vertical axis. The slope of the dose–response curve defines the extent to which a desired response alters as the dose is changed. A steeply rising and prolonged curve indicates that a small change in dose produces a large change in drug effect over a wide dose range, e.g. with the loop diuretic furosemide (used in doses from 20 mg to over 250 mg/day). By contrast, the dose–response curve for thiazide diuretics soon reaches a plateau, and the clinically useful dose range for bendroflumethiazide, for example, extends from 5 to 10 mg; increasing the dose beyond this produces no added diuretic effect, although it adds to toxicity.

Dose–response curves for wanted and unwanted effects can illustrate and quantify selective and non-selective drug action (see Fig. 8.1).

Fig. 8.1 Dose–response curves for two hypothetical drugs. For drug A, the dose that brings about the maximum wanted effect is less than the lowest dose that produces the unwanted effect. The ratio ED₅₀ (unwanted effect)/ED₅₀ (wanted effect) indicates that drug A has a large therapeutic index; it is thus highly selective in its wanted action. Drug B causes unwanted effects at doses well below producing its maximum benefit. The ratio ED₅₀ (unwanted effect)/ED₅₀ (wanted effect) indicates that the drug has a small therapeutic index: it is thus non-selective.

Potency and efficacy

A clear distinction between potency and efficacy is pertinent, particularly in relation to claims made for usefulness in therapeutics.

Potency

is the amount (weight) of drug in relation to its effect, e.g. if weight-for-weight drug A has a greater effect than drug B, then drug A is more potent than drug B, although the maximum therapeutic effect obtainable may be similar with both drugs.

The diuretic effect of bumetanide 1 mg is equivalent to that of furosemide 50 mg; thus bumetanide is more potent than furosemide but both drugs achieve about the same maximum effect. The difference in weight of drug administered is of no clinical significance unless it is great.

Pharmacological efficacy

refers to the strength of response induced by occupancy of a receptor by an agonist (intrinsic activity); it is a specialised pharmacological concept. But clinicians are concerned with therapeutic efficacy, as follows.

Therapeutic efficacy

or effectiveness, is the capacity of a drug to produce an effect and refers to the maximum such effect. For example, if drug A can produce a therapeutic effect that cannot be obtained with drug B, however much of drug B is given, then drug A has the higher therapeutic efficacy. Differences in therapeutic efficacy are of great clinical importance, usually more than potency.

Amiloride (low efficacy) can at best effect excretion of no more than 5% of the sodium load filtered by the glomeruli; there is no point in increasing the dose beyond that which achieves this, as this is its maximum diuretic effect. Bendroflumethiazide (moderate efficacy) can effect excretion of no more than 10% of the filtered sodium load no matter how large the dose. Furosemide can effect excretion of 25% and more of filtered sodium; it is a high-efficacy diuretic.

Therapeutic index

With progressive increases in dose, the desired response in the patient usually rises to a maximum beyond which further increases elicit no greater benefit but induce unwanted effects. This is because most drugs do not have a single dose–response curve, but a different curve for each action, wanted as well as unwanted. Increases in dose beyond that which gives the maximum wanted response recruit only new and unwanted actions.

A sympathomimetic bronchodilator might exhibit one dose–response relation for decreasing airway resistance (wanted) and another for increase in heart rate (unwanted). Clearly, the usefulness of any drug relates closely to the extent to which such dose–response relations overlap.

Ehrlich (see p. 162) introduced the concept of the therapeutic index or ratio as the maximum tolerated dose divided by the minimum curative dose, but the index is never calculated thus as such single doses cannot be determined accurately in humans. More realistically, a dose that has some unwanted effect in 50% of humans, e.g. in the case of an adrenoceptor agonist bronchodilator a specified increase in heart rate, is compared with that which is therapeutic in 50% (ED₅₀), e.g. a specified decrease in airways resistance.

In practice, such information is not available for many drugs but the therapeutic index does embody a concept that is fundamental in comparing the usefulness of one drug with another, namely, safety in relation to efficacy. Figure 8.1 expresses the concept diagrammatically.

Tolerance

Continuous or repeated administration of a drug is often accompanied by a gradual diminution of the effect it produces. A state of tolerance exists when it becomes necessary to increase the dose of a drug to get an effect previously obtained with a smaller dose, i.e. reduced sensitivity. By contrast, the term tachyphylaxis describes the phenomenon of progressive lessening of effect (refractoriness) in response to frequently administered doses (see Receptors, p. 75); it tends to develop more rapidly than tolerance.

The use of opioids readily illustrates tolerance, as witnessed by the huge doses of morphine that may be necessary to maintain pain relief in terminal care; the effect is due to reduced pharmacological efficacy (see above) at receptor sites or to down-regulation of receptors. Tolerance is acquired rapidly with nitrates used to prevent angina, possibly mediated by the generation of oxygen free radicals from nitric oxide; it can be avoided by removing transdermal nitrate patches for 4–8 h, e.g. at night, to allow the plasma concentration to fall.

Accelerated metabolism by enzyme induction (see p. 93) also leads to tolerance, as experience shows with alcohol, taken regularly as opposed to sporadically. There is commonly cross-tolerance between drugs of similar structure.

Failure of certain individuals to respond to normal doses of a drug, e.g. resistance to warfarin, vitamin D, constitutes a form of natural tolerance (see Pharmacogenetics, p. 101).

Bioassay and standardisation

Biological assay (bioassay) is the process by which the activity of a substance (identified or unidentified) is measured on living material: e.g. contraction of bronchial, uterine or vascular muscle. It is used only when chemical or physical methods are not practicable as in the case of a mixture of active substances, or of an incompletely purified preparation, or where no chemical method has been developed. The activity of a preparation is expressed relative to that of a standard preparation of the same substance.

Biological standardisation is a specialised form of bioassay. It involves matching of material of unknown potency with an international or national standard with the objective of providing a preparation for use in therapeutics and research. The results are expressed as units of a substance rather than its weight, e.g. insulin, vaccines.

Pharmacokinetics

To initiate a desired drug action is a qualitative choice but, when the qualitative choice is made, considerations of quantity immediately arise; it is possible to have too much or too little of a good thing. To obtain the right effect at the right intensity, at the right time, for the right duration, with minimal risk of unpleasantness or harm, is what pharmacokinetics is about.

Dosage regimens of long-established drugs grew from trial and error. Doctors learned by experience the dose, the frequency of dosing and the route of administration that was most likely to benefit and least likely to harm. But this empirical (‘suck it and see’) approach is no longer tenable. We now have an understanding of how drugs cross membranes to enter the body, how they are distributed round it in the blood and other body fluids, how they are bound to plasma proteins and tissues (which act as stores), and how they are eliminated from the body. Quantification of these processes paves the way for efficient development of dosing regimens.

Pharmacokinetics³ is concerned with the rate at which drug molecules cross cell membranes to enter the body, to distribute within it and to leave the body, as well as with the structural changes (metabolism) to which they are subject within it.

The discussion covers the following topics:

• Drug passage across cell membranes.

• Order of reaction or process (first and zero order).

• Time course of drug concentration and effect:

plasma half-life and steady-state concentration

therapeutic monitoring.

• The individual processes: absorption, distribution, metabolism (biotransformation), elimination.

Drug passage across cell membranes

Certain concepts are fundamental to understanding how drug molecules make their way around the body to achieve their effect. The first concerns the modes by which drugs cross cell membranes and cells.

Our bodies are labyrinths of fluid-filled spaces. Some, such as the lumina of the kidney tubules or intestine, connect to the outside world; the blood, lymph and cerebrospinal fluid are enclosed. Sheets of cells line these spaces, and the extent to which a drug can cross epithelia or endothelia is fundamental to its clinical use, determining whether a drug can be taken orally for systemic effect, and whether within the glomerular filtrate it will be reabsorbed or excreted in the urine.

Cell membranes are essentially bilayers of lipid molecules with ‘islands’ of protein, and they preserve and regulate the internal environment. Lipid-soluble substances diffuse readily into cells and therefore throughout body tissues. Adjacent epithelial or endothelial cells are linked by tight junctions, some of which are traversed by water-filled channels that allow the passage of water-soluble substances of small molecular size.

The jejunum and proximal renal tubule contain many such channels and are leaky epithelia, whereas the tight junctions in the stomach and urinary bladder do not have these channels and water cannot pass; they are termed tight epithelia. Special protein molecules within the lipid bilayer allow specific substances to enter or leave the cell preferentially, i.e. energy-utilising transporter processes, described later. The natural processes of passive diffusion, filtration and carrier-mediated transport determine the passage of drugs across membranes and cells, and their distribution round the body.

Passive diffusion

This is the most important means by which a drug enters the tissues and distributes through them. It refers simply to the natural tendency of any substance to move passively from an area of high concentration to one of low concentration. In the context of an individual cell, the drug moves at a rate proportional to the concentration difference across the cell membrane, i.e. it shows first-order kinetics (see p. 81); cellular energy is not required, which means that the process does not become saturated and is not inhibited by other substances.

The extent to which drugs are soluble in water or lipid is central to their capacity to cross cell membranes and depends on environmental pH and the structural properties of the molecule.

Lipid solubility is promoted by the presence of a benzene ring, a hydrocarbon chain, a steroid nucleus or halogen (-Br, -Cl, -F) groups. Water solubility is promoted by the presence of alcoholic (-OH), amide (-CO·NH₂) or carboxylic (-COOH) groups, or the formation of glucuronide and sulphate conjugates.

It is useful to classify drugs in a physicochemical sense into:

• Those that are variably ionised according to environmental pH (electrolytes) (lipid soluble or water soluble).

• Those that are incapable of becoming ionised whatever the environmental pH (un-ionised, non-polar substances) (lipid soluble).

• Those that are permanently ionised whatever the environmental pH (ionised, polar substances) (water soluble).

Drugs that ionise according to environmental pH

Many drugs are weak electrolytes, i.e. their structural groups ionise to a greater or lesser extent, according to environmental pH. Most such molecules are present partly in the ionised and partly in the un-ionised state. The degree of ionisation influences lipid solubility (and hence diffusibility) and so affects absorption, distribution and elimination.

Ionisable groups in a drug molecule tend either to lose a hydrogen ion (acidic groups) or to add a hydrogen ion (basic groups). The extent to which a molecule has this tendency to ionise is given by the dissociation (or ionisation) constant (K_a), expressed as the pK_a, i.e. the negative logarithm of the K_a (just as pH is the negative logarithm of the hydrogen ion concentration). In an acidic environment, i.e. one already containing many free hydrogen ions, an acidic group tends to retain a hydrogen ion and remains un-ionised; a relative deficit of free hydrogen ions, i.e. a basic environment, favours loss of the hydrogen ion from an acidic group, which thus becomes ionised. The opposite is the case for a base. The issue may be summarised:

• Acidic groups become less ionised in an acidic environment.

• Basic groups become less ionised in a basic (alkaline) environment and vice versa.

This in turn influences diffusibility because:

• Un-ionised drug is lipid soluble and diffusible.

• Ionised drug is lipid insoluble and non-diffusible.

Quantifying the degree of ionisation helps to express the profound effect of environmental pH. Recall that when the pH of the environment is the same as the pK_a of a drug within it, then the ratio of un-ionised to ionised molecules is 1:1. But for every unit by which pH is changed, the ratio of un-ionised to ionised molecules changes 10-fold. Thus, when the pH is 2 units less than the pK_a, molecules of an acid become 100 times more un-ionised and when the pH is 2 units more than the pK_a, molecules of an acid become 100 more ionised. Such pH change profoundly affects drug kinetics.

pH variation and drug kinetics

The pH partition hypothesis expresses the separation of a drug across a lipid membrane according to differences in environmental pH. There is a wide range of pH in the gut (pH 1.5 in the stomach, 6.8 in the upper and 7.6 in the lower intestine). But the pH inside the body is maintained within a limited range (pH 7.46 ± 0.04), so that only drugs that are substantially un-ionised at this pH will be lipid soluble, diffuse across tissue boundaries and so be widely distributed, e.g. into the central nervous system (CNS). Urine pH varies between the extremes of 4.6 and 8.2, and the prevailing pH affects the amount of drug reabsorbed from the renal tubular lumen by passive diffusion.

In the stomach, aspirin (acetylsalicylic acid, pK_a 3.5) is un-ionised and thus lipid soluble and diffusible. When aspirin enters the gastric epithelial cells (pH 7.4) it will ionise, become less diffusible and so will localise there. This ion trapping is one mechanism whereby aspirin is concentrated in, and so harms, the gastric mucosa. In the body aspirin is metabolised to salicylic acid (pK_a 3.0), which at pH 7.4 is highly ionised and thus remains in the extracellular fluid. Eventually the molecules of salicylic acid in the plasma are filtered by the glomeruli and pass into the tubular fluid, which is generally more acidic than plasma and causes a proportion of salicylic acid to become un-ionised and lipid soluble so that it diffuses back into the tubular cells. Alkalinising the urine with an intravenous infusion of sodium bicarbonate causes more salicylic acid to become ionised and lipid insoluble so that it remains in the tubular fluid, and passes into the urine. Treatment for salicylate (aspirin) overdose utilises this effect.

Conversely, acidifying the urine increases the elimination of the base amfetamine (pK_a 9.9) (see Acidification of urine, p. 126).

Drugs that are incapable of becoming ionised

These include digoxin and steroid hormones such as prednisolone. Effectively lacking any ionisable groups, they are unaffected by environmental pH, are lipid soluble and so diffuse readily across tissue boundaries. These drugs are also referred to as non-polar.

Permanently ionised drugs

Drugs that are permanently ionised contain groups that dissociate so strongly that they remain ionised over the range of the body pH. Such compounds are termed polar, for their groups are either negatively charged (acidic, e.g. heparin) or positively charged (basic, e.g. ipratropium, tubocurarine, suxamethonium) and all have a very limited capacity to cross cell membranes. This is a disadvantage with heparin, which the gut does not absorb, so that it is given parenterally. Conversely, heparin is a useful anticoagulant in pregnancy because it does not cross the placenta (which the orally effective warfarin does and is liable to cause fetal haemorrhage as well as being teratogenic).

The following are particular examples of the relevance of drug passage across membranes.

Brain and cerebrospinal fluid (CSF)

The capillaries of the cerebral circulation differ from those in most other parts of the body in that they lack the filtration channels between endothelial cells through which substances in the blood normally gain access to the extracellular fluid. Tight junctions between adjacent capillary endothelial cells, together with their basement membrane and a thin covering from the processes of astrocytes, separate the blood from the brain tissue, forming the blood–brain barrier. Compounds that are lipid insoluble do not cross it readily, e.g. atenolol, compared with propranolol (lipid soluble), and unwanted CNS effects are more prominent with the latter. Therapy with methotrexate (lipid insoluble) may fail to eliminate leukaemic deposits in the CNS.

Conversely lipid-soluble substances enter brain tissue with ease; thus diazepam (lipid soluble) given intravenously is effective within 1 min for status epilepticus, and effects of alcohol (ethanol) by mouth are noted within minutes; the level of general anaesthesia can be controlled closely by altering the concentration of inhaled anaesthetic gas (lipid soluble).

Placenta

Maternal blood bathes the chorionic villi, which consist of a layer of trophoblastic cells that enclose fetal capillaries. Their large surface area and the high placental blood flow (500 mL/min) are essential for gas exchange, uptake of nutrients and elimination of waste products. Thus a lipid barrier separates the fetal and maternal bloodstreams, allowing the passage of lipid-soluble substances but excluding water-soluble compounds, especially those with a molecular weight exceeding 600.⁴

This exclusion is of particular importance with short-term use, e.g. tubocurarine (mol. wt. 772) (lipid insoluble) or gallamine (mol. wt. 891) used as a muscle relaxant during caesarean section do not affect the infant; with prolonged use, however, all compounds will eventually enter the fetus to some extent (see Index).

Filtration

Aqueous channels in the tight junctions between adjacent epithelial cells allow the passage of some water-soluble substances. Neutral or uncharged, i.e. non-polar, molecules pass most readily because the pores are electrically charged. Within the alimentary tract, channels are largest and most numerous in jejunal epithelium, and filtration allows for rapid equilibration of concentrations and consequently of osmotic pressures across the mucosa. Ions such as sodium enter the body through the aqueous channels, the size of which probably limits passage to substances of low molecular weight, e.g. ethanol (mol. wt. 46). Filtration seems to play at most a minor role in drug transfer within the body except for glomerular filtration, which is an important mechanism of drug excretion.

Carrier-mediated transport

The membranes of many cells incorporate carrier-mediated transporter processes that control the entry and exit of endogenous molecules, and show a high degree of specificity for particular compounds because they have evolved from biological needs for the uptake of essential nutrients or elimination of metabolic products. Drugs that bear some structural resemblance to natural constituents of the body are likely to utilise these mechanisms.

Some carrier-mediated transport processes operate passively, i.e. do not require cellular energy, and this is facilitated diffusion, e.g. vitamin B₁₂ absorption. Other, energy-requiring processes move substrates into or out of cells against a concentration gradient very effectively, i.e. by active transport; they are subject to saturation, inhibition and induction (see p. 91).

The order of reaction or process

In the body, drug molecules reach their sites of action after crossing cell membranes and cells, and many are metabolised in the process. The rate at which these movements or changes take place is subject to important influences called the order of reaction or process. In biology generally, two orders of such reactions are recognised, and are summarised as follows:

• First-order processes by which a constant fraction of drug is transported/metabolised in unit time.

• Zero-order processes by which a constant amount of drug is transported/metabolised in unit time.

First-order (exponential) processes

In the majority of instances, the rates at which absorption, distribution, metabolism and excretion of a drug occur are directly proportional to its concentration in the body. In other words, transfer of drug across a cell membrane or formation of a metabolite is high at high concentrations and falls in direct proportion to be low at low concentrations (an exponential relationship).

This is because the processes follow the Law of Mass Action, which states that the rate of reaction is directly proportional to the active filtration masses of reacting substances. In other words, at high concentrations there are more opportunities for crowded molecules to interact with one another or to cross cell membranes than at low, uncrowded concentrations. Processes for which the rate of reaction is proportional to the concentration of participating molecules are first-order processes.

In doses used clinically, most drugs are subject to first-order processes of absorption, distribution, metabolism and elimination, and this knowledge is useful. The current chapter later describes how the rate of elimination of a drug from the plasma falls as the concentration in plasma falls, and the time for any plasma concentration to fall by 50% (t_½, the plasma half-life) is always the same. Thus, it becomes possible to quote a constant value for the t_½ of the drug. This occurs because rate and concentration are in proportion, i.e. the process obeys first-order kinetics.

Knowing that first-order conditions apply to a drug allows accurate calculations that depend on its t_½, i.e. time to achieve steady-state plasma concentration, time to elimination, and the construction of dosing schedules.

Zero-order processes (saturation kinetics)

As the amount of drug in the body rises, metabolic reactions or processes that have limited capacity become saturated. In other words, the rate of the process reaches a maximum amount at which it stays constant, e.g. due to limited activity of an enzyme, and any further increase in rate is impossible despite an increase in the dose of drug. In these circumstances, the rate of reaction is no longer proportional to dose, and exhibits rate-limited or dose-dependent⁵ or zero-order or saturation kinetics. In practice, enzyme-mediated metabolic reactions are the most likely to show rate limitation because the amount of enzyme present is finite and can become saturated. Passive diffusion does not become saturated. There are some important consequences of zero-order kinetics.

Alcohol

Alcohol is subject to first-order kinetics with a t_½ of about 1 h at plasma concentrations below 10 mg/dL (attained after drinking about two-thirds of a unit (glass) of wine or beer). Above this concentration the main enzyme (alcohol dehydrogenase) that converts the alcohol into acetaldehyde approaches and then reaches saturation, at which point alcohol metabolism cannot proceed any faster than about 10 mL or 8 g/h for a 70-kg man. If the subject continues to drink, the blood alcohol concentration rises disproportionately, for the rate of metabolism remains the same, as alcohol shows zero-order kinetics.

An illustration. Consider a man of average size who drinks about half (375 mL) a standard bottle of whisky (40% alcohol), i.e. 150 mL alcohol, over a short period, absorbs it and goes drunk to bed at midnight with a blood alcohol concentration of about 250 mg/dL. If alcohol metabolism were subject to first-order kinetics, with a t_½ of 1 h throughout the whole range of social consumption, the subject would halve his blood alcohol concentration each hour (see Fig. 8.2). It is easy to calculate that, when he drives his car to work at 08.00 hours the next morning, he has a negligible blood alcohol concentration (less than 1 mg/dL) though, no doubt, a hangover might reduce his driving skill.

Fig. 8.2 Changes in plasma concentration following an intravenous bolus injection of a drug in the elimination phase (the distribution phase, see text, is not shown). As elimination is a first-order process, the time for any concentration point to fall by 50% (t_½) is always the same.

But at these high concentrations, alcohol is in fact subject to zero-order kinetics and so, metabolising about 10 mL alcohol per hour, after 8 h the subject has eliminated only 80 mL, leaving 70 mL in his body and giving a blood concentration of about 120 mg/dL. At this level, his driving skill is seriously impaired. The subject has an accident on his way to work and is breathalysed despite his indignant protests that he last touched a drop before midnight. Banned from the road, on his train journey to work he will have leisure to reflect on the difference between first-order and zero-order kinetics (though this is unlikely!).

In practice. The example above describes an imagined event but similar cases occur in everyday therapeutics. Phenytoin, at low dose, exhibits a first-order elimination process and there is a directly proportional increase in the steady-state plasma concentration with increase in dose. But gradually the enzymatic elimination process approaches and reaches saturation, the process becoming constant and zero order. While the dosing rate can be increased, the metabolism rate cannot, and the plasma concentration rises steeply and disproportionately, with danger of toxicity. Salicylate metabolism also exhibits saturation kinetics but at high therapeutic doses. Clearly saturation kinetics is a significant factor in delay of recovery from drug overdose, e.g. with aspirin or phenytoin.

Order of reaction and t_½. When a drug is subject to first-order kinetics, the t_½ is a constant characteristic, i.e. a constant value can be quoted throughout the plasma concentration range (accepting that there will be variation in t_½ between individuals), and this is convenient. But if the rate of a process is not directly proportional to plasma concentration, then the t_½ cannot be constant. Consequently, no single value for t_½ describes overall elimination when a drug exhibits zero-order kinetics. In fact, t_½ decreases as plasma concentration falls and the calculations on elimination and dosing that are so easy with first-order elimination (see below) become more complicated.

Zero-order absorption processes apply to iron, to depot intramuscular formulations and to drug implants, e.g. antipsychotics and sex hormones.

Time course of drug concentration and effect

Plasma half-life and steady-state concentration

The manner in which plasma drug concentration rises or falls when dosing begins, alters or ceases follows certain simple rules, which provide a means for rational control of drug effect. Central to understanding these is the concept of half-life (t_½) or half-time.

Decrease in plasma concentration after an intravenous bolus injection

Following an intravenous bolus injection (a single dose injected in a period of seconds as distinct from a continuous infusion), plasma concentration rises quickly as drug enters the blood to reach a peak. There is then a sharp drop as the drug distributes round the body (distribution phase), followed by a steady decline as drug is removed from the blood by the liver or kidneys (elimination phase). If the elimination processes are first order, the time taken for any concentration point in the elimination phase to fall to half its value (the t_½) is always the same; see Figure 8.2. Note that the drug is virtually eliminated from the plasma in five t_½ periods.

The t_½ is the one pharmacokinetic value of a drug that it is most useful to know.

Increase in plasma concentration with constant dosing

With a constant rate infusion, the amount of drug in the body and with it the plasma concentration rise until a state is reached at which the rate of administration to the body is exactly equal to the rate of elimination from it: this is called the steady state. The plasma concentration is then on a plateau, and the drug effect is stable. Figure 8.3 depicts the smooth changes in plasma concentration that result from a constant intravenous infusion. Clearly, giving a drug by regularly spaced oral or intravenous doses will result in plasma concentrations that fluctuate between peaks and troughs, but in time all of the peaks will be of equal height and all of the troughs will be of equal depth; this is also called a steady-state concentration, as the mean concentration is constant.⁶

Fig. 8.3 Changes in plasma concentration during the course of a constant-rate intravenous infusion. (a) The infusion commences and plasma concentration rises to reach a steady state (plateau) in about 5 × t_½ periods. (b) The infusion rate is increased by 50% and the plasma concentration rises further to reach a new steady state that is 50% higher than the original steady state; the process takes another 5 × t_½ periods. (c) The infusion is decreased to the original rate and the plasma concentration returns to the original steady state in 5 × t_½ periods. (d) The infusion is discontinued and the plasma concentration falls to virtually zero in 5 × t_½ periods.

Time to reach steady state

It is important to know when a drug administered at a constant rate achieves a steady-state plasma concentration, for maintaining the same dosing schedule then ensures a constant amount of drug in the body and the patient will experience neither acute toxicity nor decline of effect. The t_½ provides the answer. Taking ultimate steady state attained as 100%:

in 1 × t_½ the concentration will be (100/2) 50%

in 2 × t_½ (50 + 50/2) 75%

in 3 × t_½ (75 + 25/2) 87.5%

in 4 × t_½ (87.5 + 2.5/2) 93.75%

in 5 × t_½ (93.75 + 6.25/2) 96.875%

of the ultimate steady state.

When a drug is given at a constant rate (continuous or repeated administration), the time to reach steady state depends only on the t_½ and, for all practical purposes, after 5 × t_½ periods the amount of drug in the body is constant and the plasma concentration is at a plateau (a–b in Fig. 8.3).

Change in plasma concentration with change or cessation of dosing

The same principle holds for change from any steady-state plasma concentration to a new steady state brought about by increase or decrease in the rate of drug administration. Provided the kinetics remain first order, increasing or decreasing the rate of drug administration (b and c in Fig. 8.3) gives rise to a new steady-state concentration in a time equal to 5 × t_½ periods.

Similarly, starting at any steady-state plasma concentration (100%), discontinuing the dose (d in Fig. 8.3) will cause the plasma concentration to fall to virtually zero in 5 × t_½ periods, as described in Figure 8.2.

Note that the difference between the rate of drug administration (input) and the rate of elimination (output) determines the actual level of any steady-state plasma concentration (as opposed to the time taken to reach it). If drug elimination remains constant and administration increases by 50%, in time the plasma concentration will reach a new steady-state concentration, which will be 50% greater than the original.

The relation between t_½ and time to reach steady-state plasma concentration applies to all drugs that obey first-order kinetics. This holds as much to dobutamine (t_½ 2 min), when it is useful to know that an alteration of infusion rate will reach a plateau within 10 min, as to digoxin (t_½ 36 h), when a constant daily oral dose will give a steady-state plasma concentration only after 7.5 days. This book quotes plasma t_½ values where they are relevant. Inevitably, natural variation within the population produces a range in t_½ values for any drug and the text quotes only single average t_½ values while recognising that the population range may be as much as 50% from the stated figure in either direction.

Some t_½ values appear in Table 8.1 to illustrate their range and implications for dosing in clinical practice.

Table 8.1 Plasma t_½ of some drugs

Drug	t_½
Adenosine	< 2 s
Dobutamine	2 min
Benzylpenicillin	30 min
Amoxicillin	1 h
Paracetamol	2 h
Midazolam	3 h
Tolbutamide	6 h
Atenolol	7 h
Dosulepin	25 h
Diazepam	40 h
Piroxicam	45 h
Ethosuximide	54 h

Biological effect t_½

is the time in which the biological effect of a drug declines by one-half. With drugs that act competitively on receptors (α- and β-adrenoceptor agonists and antagonists) the biological effect t_½ can be estimated with reasonable accuracy. Sometimes the biological effect t_½ cannot be provided, e.g. with antimicrobials when the number of infecting organisms and their sensitivity determine the outcome.

Therapeutic drug monitoring

Patients differ greatly in the dose of drug required to achieve the same response. The dose of warfarin that maintains a therapeutic concentration may vary as much as five-fold between individuals. This is a consequence of variation in rates of drug metabolism, disposition and tissue responsiveness, and it raises the question of how optimal drug effect can be achieved quickly for the individual patient.

In principle, drug effect relates to free (unbound) concentration at the tissue receptor site, which in turn reflects (but is not necessarily the same as) the concentration in the plasma. For many drugs, correlation between plasma concentration and effect is indeed better than that between dose and effect. Yet monitoring therapy by measuring drug in plasma is of practical use only in selected instances. The underlying reasons repay some thought.

Plasma concentration may not be worth measuring

where dose can be titrated against a quickly and easily measured effect such as blood pressure (antihypertensives), body-weight (diuretics), INR (oral anticoagulants) or blood sugar (hypoglycaemics).

Plasma concentration has no correlation with effect

with drugs that act irreversibly (named ‘hit and run drugs’ because their effect persists long after the drug has left the plasma). Such drugs inactivate targets (enzyme, receptor) and restoration of effect occurs only after days or weeks, when resynthesis takes place, e.g. some monoamine oxidase inhibitors, aspirin (on platelets), some anticholinesterases and anticancer drugs.

Plasma concentration may correlate poorly with effect

When a drug is metabolised to several products, active to varying degree or inactive, the assay of the parent drug alone is unlikely to reflect its activity, e.g. some benzodiazepines. Similarly binding of basic drugs, e.g. lidocaine, to acute phase proteins, e.g. α₁-acid glycoprotein, spuriously increases the total concentration in plasma. The best correlation is likely to be achieved by measurement of free (active) drug in plasma water, but this is technically more difficult and total drug in plasma is usually monitored in routine clinical practice. Saliva is sometimes used.

Plasma concentration may correlate well with effect

Plasma concentration monitoring has proved useful:

• As a guide to the effectiveness of therapy, e.g. plasma gentamicin and other antimicrobials against sensitive bacteria, plasma theophylline for asthma, plasma ciclosporin to avoid transplant rejection, lithium for mood disorder.

• To reduce the risk of adverse drug effects when therapeutic doses are close to toxic doses (low therapeutic index), e.g. otic damage with aminoglycoside antibiotics; adverse CNS effects of lithium, nephrotoxicity with ciclosporin.

• When the desired effect is suppression of infrequent sporadic events such as epileptic seizures or episodes of cardiac arrhythmia.

• To check patient compliance on a drug regimen, when there is failure of therapeutic effect at a known effective dose, e.g. antiepilepsy drugs.

• To diagnose and manage drug overdose.

• When lack of therapeutic effect and toxicity may be difficult to distinguish. Digoxin is both a treatment for, and sometimes the cause of, cardiac supraventricular tachycardia; a plasma digoxin measurement will help to distinguish whether an arrhythmia is due to too little or too much digoxin.

Interpreting plasma concentration measurements

• The target therapeutic concentration range for a drug is a guide to optimise dosing together with other clinical indicators of progress.

• Take account of the time needed to reach steady-state dosing conditions (see above). Additionally, some drugs alter their own rates of metabolism by enzyme induction, e.g. carbamazepine and phenytoin, and it is best to allow 2–4 weeks between change in dose and meaningful plasma concentration measurement.

• As a general rule, when a drug has a short t_½ it is desirable to know both peak (15 min after an intravenous dose) and trough (just before the next dose) concentrations to provide efficacy without toxicity, as with gentamicin (t_½ 2.5 h). For a drug with a long t_½, it is usually best to sample just before a dose is due; effective immunosuppression with ciclosporin (t_½ 27 h) is obtained with trough concentrations of 50–200 micrograms/L when the drug is given by mouth.

Individual pharmacokinetic processes

Drug absorption into, distribution around, metabolism by and elimination from the body are reviewed.

Absorption

Commonsense considerations of anatomy, physiology, pathology, pharmacology, therapeutics and convenience determine the routes by which drugs are administered. Usually these are:

• Enteral: by mouth (swallowed) or by sublingual or buccal absorption; by rectum.

• Parenteral: by intravenous injection or infusion, intramuscular injection, subcutaneous injection or infusion, inhalation, topical application for local (skin, eye, lung) or for systemic (transdermal) effect.

• Other routes, e.g. intrathecal, intradermal, intranasal, intratracheal, intrapleural, are used when appropriate.

The features of the various routes, their advantages and disadvantages are relevant.

Absorption from the gastrointestinal tract

The small intestine is the principal site for absorption of nutrients and it is also where most orally administered drugs enter the body. This part of the gut has an enormous surface area due to the intestinal villi, and an epithelium through which fluid readily filters in response to osmotic differences caused by the presence of food. Disturbed alimentary motility can reduce drug absorption, i.e. if food slows gastric emptying, or gut infection accelerates intestinal transit. Additionally, it is becoming apparent that uptake and efflux transporters in enterocytes (see p. 93) play a substantial role in controlling the absorption of certain drugs, e.g. digoxin, ciclosporin. Many sustained-release formulations probably depend on absorption from the colon.

Absorption of ionisable drugs from the buccal mucosa responds to the prevailing pH, which is 6.2–7.2. Lipid-soluble drugs are rapidly effective by this route because blood flow through the mucosa is abundant; these drugs enter directly into the systemic circulation, avoiding the possibility of first-pass (presystemic) inactivation by the liver and gut (see below).

The stomach does not play a major role in absorbing drugs, even those that are acidic and thus un-ionised and lipid soluble at gastric pH, because its surface area is much smaller than that of the small intestine and gastric emptying is speedy (t_½ 30 min).

Enterohepatic circulation

This system is illustrated by the bile salts which are formed in the liver, then conserved by circulating round liver, intestine and portal blood about eight times a day. Several drugs form conjugates with glucuronic acid in the liver and enter the bile. Too polar (ionised) to be reabsorbed, the glucuronides remain in the gut, are hydrolysed by intestinal enzymes and bacteria, and the parent drug, thus released, is reabsorbed and reconjugated in the liver. Enterohepatic recycling appears to help sustain the plasma concentration and so the effect of sulindac, pentaerithrityl tetranitrate and ethinylestradiol (in many oral contraceptives).

Systemic availability and bioavailability

A drug injected intravenously enters the systemic circulation and thence gains access to the tissues and to receptors, i.e. 100% is available to exert its therapeutic effect. If the same quantity of the drug is swallowed, it does not follow that the entire amount will reach first the portal blood and then the systemic blood, i.e. its availability for therapeutic effect via the systemic circulation may be less than 100%. The anticipated response to a drug must take account of its availability to the systemic circulation.

While considerations of reduced availability attach to any drug given by any route other than intravenously, and intended for systemic effect, in practice the issue concerns enteral administration. The extent of systemic availability is usually calculated by relating the area under the plasma concentration–time curve (AUC) after a single oral dose to that obtained after intravenous administration of the same amount (by which route a drug is 100% systemically available). Calculation of AUCs after oral doses also allows a comparison of the bioavailability of different pharmaceutical formulations of the same drug. Factors influencing systemic availability present in three main ways, as described below.

Pharmaceutical factors ⁷

The amount of drug released from a dose form (and so becoming available for absorption) is referred to as its bioavailability. This is highly dependent on its pharmaceutical formulation. With tablets, for example, particle size (surface area exposed to solution), diluting substances, tablet size and pressure used in the tabletting process can affect disintegration and dissolution and so the bioavailability of the drug. Manufacturers must test their products to ensure that their formulations release the same amount of drug at the same speed from whatever manufactured batch or brand the patient may be taking.

Differences in bioavailability are prone to occur with modified-release (m/r) formulations, i.e. where the rate or place of release of the active ingredients has been modified (also called sustained, controlled or delayed release) (see p. 97). Modified-release preparations from different manufacturers may differ in their bioavailability profiles despite containing the same amount of drug, i.e. there is neither bioequivalence nor therapeutic equivalence, and the problem is particularly acute where the therapeutic ratio is narrow. In this case, ‘brand name prescribing’, i.e. using only a particular brand name for a particular patient is justified, e.g. for m/r preparations of theophylline, lithium, nifedipine and diltiazem.

Physicians tend to ignore pharmaceutical formulation as a factor in variable or unexpected responses because they do not understand it and feel entitled to rely on reputable manufacturers and official regulatory authorities to ensure provision of reliable formulations. Good pharmaceutical companies reasonably point out that, having a reputation to lose, they take much trouble to make their preparations consistently reliable. This is a matter of great importance when dosage must be precise (anticoagulants, antidiabetics, adrenal corticosteroids).

Biological factors

Biological factors related to the gut include limitation of drug absorption by drug transporter systems (see p. 93), destruction of drug by gastric acid, e.g. benzylpenicillin, and impaired absorption due to rapid intestinal transit, which is important for all drugs that are absorbed slowly. Drugs may also bind to food constituents, e.g. tetracyclines to calcium (in milk), and to iron, or to other drugs (e.g. acidic drugs to colestyramine), and the resulting complex is not absorbed.

Presystemic (first-pass) elimination

Some drugs readily enter gut mucosal cells, but appear in low concentration in the systemic circulation. The reason lies in the considerable extent to which such drugs are metabolised in a single passage through the gut mucosa and (principally) the liver. As little as 10–20% of the parent drug may reach the systemic circulation unchanged. By contrast, after intravenous administration, 100% becomes systemically available and the patient experiences higher concentrations with greater, but more predictable, effect. Dosing, particularly initial doses, must take account of discrepancy in anticipated plasma concentrations between the intravenous and oral routes. The difference is usually less if a drug produces active metabolites.

Once a drug is in the systemic circulation, irrespective of which route is used, about 20% is subject to the hepatic metabolic processes in each circulation time because that proportion of cardiac output passes to the liver.

As the degree of presystemic elimination differs much between drugs and individuals, the phenomenon of first-pass elimination adds to variation in systemic plasma concentrations, and thus particularly in initial response to the drugs that are subject to this process. In drug overdose, decreased presystemic elimination with increased bioavailability may account for the rapid onset of toxicity with antipsychotic drugs.

Drugs for which presystemic elimination is significant include:⁸

Analgesics	Adrenoceptor blockers	Others
morphine	labetalol	chlorpromazine
	propranolol	isosorbide dinitrate
	metoprolol	nortriptyline

In severe hepatic cirrhosis with both impaired liver cell function and well-developed vessels shunting blood into the systemic circulation without passing through the liver, first-pass elimination reduces and systemic availability is increased. The result of these changes is an increased likelihood of exaggerated response to normal doses of drugs having high hepatic clearance and, on occasion, frank toxicity.

Drugs that exhibit the hepatic first-pass phenomenon do so because of the rapidity with which they are metabolised. The rate of delivery to the liver, i.e. blood flow, is then the main determinant of its rate of metabolism. Many other drugs are completely metabolised by the liver but at a slower rate and consequently loss in the first pass through the liver is unimportant. Dose adjustment to account for presystemic elimination is unnecessary, e.g. for diazepam, phenytoin, theophylline, warfarin.

Advantages and disadvantages of enteral administration

By swallowing

For systemic effect

Advantages are convenience and acceptability.

Disadvantages are that absorption may be delayed, reduced or even enhanced after food, or slow or irregular after drugs that inhibit gut motility (antimuscarinic, opioid). Differences in presystemic elimination are a cause of variation in drug effect between patients. Some drugs are not absorbed (gentamicin) and others are destroyed in the gut (insulin, oxytocin, some penicillins). Tablets taken with too small a quantity of liquid and in the supine position, can lodge in the oesophagus with delayed absorption ⁹ and may even cause ulceration (sustained-release potassium chloride and doxycycline tablets), especially in the elderly and those with an enlarged left atrium which impinges on the oesophagus.¹⁰

For effect in the gut

Advantages are that the drug is placed at the site of action (neomycin, anthelminthics), and with non-absorbed drugs the local concentration can be higher than would be safe in the blood.

Disadvantages are that drug distribution may be uneven, and in some diseases of the gut the whole thickness of the wall is affected (severe bacillary dysentery, typhoid) and effective blood concentrations (as well as luminal concentrations) may be needed.

Sublingual or buccal for systemic effect

Advantages are that the effect is quick, e.g. with glyceryl trinitrate as an aerosol spray, or as sublingual tablets that are chewed, giving greater surface area for solution. Spitting out the tablet will terminate the effect.

Disadvantages are the inconvenience if use has to be frequent, irritation of the mucous membrane and excessive salivation, which promotes swallowing, so losing the advantages of bypassing presystemic elimination.

Rectal administration

For systemic effect (suppositories or solutions)

The rectal mucosa has a rich blood and lymph supply and, in general, dose requirements are either the same or slightly greater than those needed for oral use. Drugs chiefly enter the portal system, but those that are subject to hepatic first-pass elimination may escape this if they are absorbed from the lower rectum, which drains directly to the systemic circulation. The degree of presystemic elimination thus depends on distribution within the rectum and this is somewhat unpredictable.

Advantages are that a suppository can replace a drug that irritates the stomach (aminophylline, indometacin); the route is suitable in vomiting, motion sickness, migraine or when a patient cannot swallow, and when cooperation is lacking (sedation in children).

Disadvantages are psychological in that the patient may be embarrassed or may even like the route too much; rectal inflammation may occur with repeated use and absorption can be unreliable, especially if the rectum is full of faeces.

For local effect,

e.g. in proctitis or colitis, is an obvious use. A survey in the UK showed that a substantial proportion of patients did not remove the wrapper before inserting the suppository!

Advantages and disadvantages of parenteral administration

(for systemic and local effect)

Intravenous (bolus or infusion)

An intravenous bolus, i.e. rapid injection, passes round the circulation being progressively diluted each time; it is delivered principally to the organs with high blood flow (brain, liver, heart, lung, kidneys).

Advantages are that the intravenous route gives swift, effective and highly predictable blood concentration and allows rapid modification of dose, i.e. immediate cessation of administration is possible if unwanted effects occur during administration. The route is suitable for administration of drugs that are not absorbed from the gut or are too irritant (anticancer agents) to be given by other routes.

Disadvantages are the hazard if drug administration is too rapid, as plasma concentration may rise at a rate such that normal mechanisms of distribution and elimination are outpaced. Some drugs will act within one arm-to-tongue (brain) circulation time, which is 13 ± 3 s; with most drugs an injection given over four or five circulation times seems sufficient to avoid excessive plasma concentrations. Local venous thrombosis is liable to occur with prolonged infusion and with bolus doses of irritant formulations, e.g. diazepam, or microparticulate components of infusion fluids, especially if small veins are used, and extravasation may be very painful and damaging. Infection of the intravenous catheter and the small thrombi on its tip is also a risk during prolonged infusions.

Intramuscular injection

Blood flow is greater in the muscles of the upper arm than in the gluteal mass and thigh, and increases with physical exercise.

Advantages are that the route is reliable, suitable for some irritant drugs, and depot preparations (neuroleptics, hormonal contraceptives) are suitable for administration at monthly or longer intervals. Absorption is more rapid than following subcutaneous injection (soluble preparations are absorbed within 10–30 min).

Disadvantages are that the route is not acceptable for self-administration, it may be painful, and if any adverse effects occur with a depot formulation, it may not be removable.

Subcutaneous injection

Advantages are that the route is reliable and is acceptable for self-administration.

Disadvantages are poor absorption in peripheral circulatory failure. Repeated injections at one site can cause lipoatrophy, resulting in erratic absorption (see Insulin, Ch. 36).

By inhalation

As a gas,

e.g. volatile anaesthetics.

As an aerosol,

e.g. β₂-adrenoceptor agonist bronchodilators. Aerosols are particles dispersed in a gas, the particles being small enough to remain in suspension for a long time instead of sedimenting rapidly under the influence of gravity; the particles may be liquid (fog) or solid (smoke). A nebuliser allows larger doses (see p. 478).

As a powder,

e.g. sodium cromoglicate. Particle size and air-flow velocity are important. Most particles greater than 5 micrometres in diameter impact in the upper respiratory areas; particles of about 2 micrometres reach the terminal bronchioles; a large proportion of particles less than 1 micrometre are exhaled. Air-flow velocity diminishes considerably as the bronchi progressively divide, promoting drug deposition peripherally.

Advantages are the rapid uptake or elimination of drugs as gases, giving the close control that has marked the use of this route in general anaesthesia from its earliest days. Self-administration is practicable. Aerosols and powders provide high local concentration for action on bronchi, minimising systemic effects.

Disadvantages are that special apparatus is needed (some patients find pressurised aerosols difficult to use to best effect) and a drug must be non-irritant if the patient is conscious. Obstructed bronchi (mucus plugs in asthma) may cause therapy to fail.

Topical application

For local effect,

e.g. to skin, eye, lung, anal canal, rectum, vagina.

Advantage is the provision of high local concentration without systemic effect (usually ¹¹).

Disadvantage is that absorption can occur, especially when there is tissue destruction so that systemic effects result, e.g. adrenal corticosteroids and neomycin to the skin, atropine to the eye. Ocular administration of a β-adrenoceptor blocker may cause systemic effects (bypassing first-pass elimination) and such eye drops are contraindicated in asthma or chronic lung disease.¹² There is extensive literature on this subject characterised by expressions of astonishment that serious effects, even death, can occur.

For systemic effect

Transdermal delivery systems release drug through a rate-controlling membrane into the skin and so into the systemic circulation. This avoids the fluctuations in plasma concentration associated with other routes of administration, as is first-pass elimination in the liver. Glyceryl trinitrate and postmenopausal hormone replacement therapy are available in the form of a sticking plaster attached to the skin or as an ointment (glyceryl trinitrate). One treatment for migraine is a nasal spray containing sumatriptan.

Distribution

If a drug is required to act throughout the body or to reach an organ inaccessible to topical administration, it must get into the blood and other body compartments. Most drugs distribute widely, in part dissolved in body water, in part bound to plasma proteins, in part to tissues. Distribution is often uneven, for drugs may bind selectively to plasma or tissue proteins or be localised within particular organs. Clearly, the site of localisation of a drug is likely to influence its action, e.g. whether it crosses the blood–brain barrier to enter the brain; the extent (amount) and strength (tenacity) of protein or tissue binding (stored drug) will affect the time it spends in the body and thereby its duration of action.

Distribution volume

The pattern of distribution from plasma to other body fluids and tissues is a characteristic of each drug that enters the circulation, and it varies between drugs. Precise information on the concentration of drug attained in various tissues and fluids is usually not available for humans.¹³ But blood plasma is sampled readily in humans, the drug concentration in which, taking account of the dose given, is a measure of whether a drug tends to remain in the circulation or to distribute from the plasma into the tissues. In other words:

• If a drug remains mostly in the plasma, its distribution volume will be small.

• If a drug is present mainly in other tissues, the distribution volume will be large.

Such information can be useful. In drug overdose, if a major proportion of the total body load is known to be in the plasma, i.e. the distribution volume is small, then haemodialysis/filtration is likely to be a useful option (as is the case with severe salicylate poisoning), but it is an inappropriate treatment for overdose with dosulepin (see Table 8.2).

Table 8.2 Apparent distribution volume of some drugs (values are in litres for a 70-kg person who would displace about 70 L)*

The principle for measuring the distribution volume is essentially that of using a dye to find the volume of a container filled with liquid. The weight of added dye divided by the concentration of dye once mixing is complete gives the distribution volume of the dye, which is the volume of the container. Similarly, the distribution volume of a drug in the body may be determined after a single intravenous bolus dose by dividing the dose given by the concentration achieved in plasma.¹⁴

The result of this calculation, the distribution volume, in fact only rarely corresponds with a physiological body space such as extracellular water or total body water, for it is a measure of the volume a drug would apparently occupy knowing the dose given and the plasma concentration achieved, and assuming the entire volume is at that concentration. For this reason, the term apparent distribution volume is often preferred. Indeed, the apparent distribution volume of some drugs that bind extensively to extravascular tissues, which is based on the resulting low plasma concentration, is many times total body volume.

The distribution volume of a drug is the volume in which it appears to distribute (or which it would require) if the concentration throughout the body were equal to that in plasma, i.e. as if the body were a single compartment.

The list in Table 8.2 illustrates a range of apparent distribution volumes. The names of those substances that distribute within (and have been used to measure) physiological spaces are printed in italics.

Selective distribution

within the body occurs because of special affinity between particular drugs and particular body constituents. Many drugs bind to proteins in the plasma; phenothiazines and chloroquine bind to melanin-containing tissues, including the retina, which may explain the occurrence of retinopathy. Drugs may also concentrate selectively in a particular tissue because of specialised transport mechanisms, e.g. iodine in the thyroid.

Plasma protein and tissue binding

Many natural substances circulate around the body partly free in plasma water and partly bound to plasma proteins; these include cortisol, thyroxine, iron, copper and, in hepatic or renal failure, by-products of physiological intermediary metabolism.

Drugs, too, circulate in the protein-bound and free states, and the significance is that the free fraction is pharmacologically active whereas the protein-bound component is a reservoir of drug that is inactive because of this binding. Free and bound fractions are in equilibrium, and free drug removed from the plasma by metabolism, renal function or dialysis is replaced by drug released from the bound fraction.

Albumin

is the main binding protein for many natural substances and drugs. Its complex structure has a net negative charge at blood pH and a high capacity but low (weak) affinity for many basic drugs, i.e. a lot is bound but it is readily released. Two particular sites on the albumin molecule bind acidic drugs with high affinity (strongly), but these sites have low capacity. Saturation of binding sites on plasma proteins in general is unlikely in the doses in which most drugs are used.

Other binding proteins in the blood include lipoprotein and α₁-acid glycoprotein, both of which carry basic drugs such as quinidine, chlorpromazine and imipramine. Thyroxine and sex hormones are bound in the plasma to specific globulins.

Disease

may modify protein binding of drugs to an extent that is clinically relevant, as Table 8.3 shows. In chronic renal failure, hypoalbuminaemia and retention of products of metabolism that compete for binding sites on protein are both responsible for the decrease in protein binding of drugs. Most affected are acidic drugs that are highly protein bound, e.g. phenytoin, and initiating or modifying the dose of such drugs for patients with renal failure requires special attention (see also Prescribing in renal disease, p. 462).

Table 8.3 Examples of plasma protein binding of drugs and effects of disease

Drug	% Unbound (free)
Warfarin	1
Diazepam	2 (6% in liver disease)
Furosemide	2 (6% in nephrotic syndrome)
Tolbutamide	2
Amitriptyline	5
Phenytoin	9 (19% in renal disease)
Triamterene	19 (40% in renal disease)
Trimethoprim	30
Theophylline	35 (71% in liver disease)
Morphine	65
Digoxin	75 (82% in renal disease)
Amoxicillin	82
Ethosuximide	100

Chronic liver disease also leads to hypoalbuminaemia and an increase of endogenous substances such as bilirubin that may compete for binding sites on protein. Drugs that are normally extensively protein bound should be used with special caution, for increased free concentration of diazepam, tolbutamide and phenytoin have been demonstrated in patients with this condition (see also Prescribing for patients with liver disease, p. 547).

The free, unbound, and therefore pharmacologically active percentages of some drugs appear in Table 8.3 to illustrate the range and, in some cases, changes recorded in disease.

Tissue binding

Some drugs distribute readily to regions of the body other than plasma, as a glance at Table 8.2 will show. These include many lipid-soluble drugs, which may enter fat stores, e.g. most benzodiazepines, verapamil and lidocaine. There is less information about other tissues, e.g. muscle, than about plasma protein binding because solid tissue samples require invasive biopsy. Extensive binding to tissues delays elimination from the body and accounts for the long t_½ of chloroquine and amiodarone.

Metabolism

The body treats most drugs as foreign substances (xenobiotics) and subjects them to various mechanisms for eliminating chemical intruders.

Metabolism is a general term for chemical transformations that occur within the body and its processes change drugs in two major ways by:

• reducing lipid solubility

• altering biological activity.

Reducing lipid solubility

Metabolic reactions tend to make a drug molecule progressively more water soluble and so favour its elimination in the urine.

Drug-metabolising enzymes developed during evolution to enable the body to dispose of lipid-soluble substances such as hydrocarbons, steroids and alkaloids that are ingested with food. Some environmental chemicals may persist indefinitely in our fat deposits, e.g. dicophane (DDT), with consequences that are currently unknown.

Altering biological activity

The end-result of metabolism usually is the abolition of biological activity, but various steps in between may have the following consequences:

1. Conversion of a pharmacologically active to an inactive substance – this applies to most drugs.

2. Conversion of one pharmacologically active to another active substance – this has the effect of prolonging drug action, as shown below.

Active drug	Active metabolite
amitriptyline	nortriptyline
codeine	morphine
chloroquine	hydroxychloroquine
diazepam	oxazepam
spironolactone	canrenone

3. Conversion of a pharmacologically inactive to an active substance (then called a prodrug). The process then follows 1 or 2, above.

Inactive substance	Active metabolite(s)	Comment
aciclovir	aciclovir triphosphate	see p. 213
colecalciferol	calcitriol and alfacalcidol	highly active metabolites of vitamin D₃; see p. 635
cyclophosphamide	phosphoramide mustard	another metabolite, acrolein, causes the bladder toxicity; see p. 517
perindopril	perindoprilat	less risk of first dose hypotension (applies to all ACE inhibitors except captopril)
levodopa	dopamine	levodopa, but not dopamine, can cross the blood–brain barrier
sulindac	sulindac sulphide	possibly reduced gastric toxicity
sulfasalazine	5-aminosalicylic acid	see p. 541
zidovudine	zidovudine triphosphate	see p. 217

The metabolic processes

The liver is by far the most important drug-metabolising organ, although a number of tissues, including the kidney, gut mucosa, lung and skin, also contribute. It is useful to think of drug metabolism in two broad phases.

Phase I

metabolism brings about a change in the drug molecule by oxidation, reduction or hydrolysis and usually introduces or exposes a chemically active site on it. The new metabolite often has reduced biological activity and different pharmacokinetic properties, e.g. a shorter t_½.

The principal group of reactions is the oxidations, in particular those undertaken by the (microsomal) mixed-function oxidases which, as the name indicates, are capable of metabolising a wide variety of compounds. The most important of these is a large ‘superfamily’ of haem proteins, the cytochrome P450 enzymes, which metabolise chemicals from the environment, the diet and drugs. By a complex process, the drug molecule incorporates one atom of molecular oxygen (O₂) to form a (chemically active) hydroxyl group and the other oxygen atom converts to water.

The following explanation provides a background to the P450 nomenclature that accompanies accounts of the metabolism of several individual drugs in this book. The many cytochrome P450 isoenzymes ¹⁵ are indicated by the letters CYP (from cytochrome P450) followed by a number denoting a family group, then a subfamily letter, and then a number for the individual enzyme within the family: for example, CYP2E1 is an isoenzyme that catalyses a reaction involved in the metabolism of alcohol, paracetamol, estradiol and ethinylestradiol.

The enzymes of families CYP1, 2 and 3 metabolise 70–80% of clinically used drugs as well as many other foreign chemicals and, within these, CYP3A, CYP2D and CYP2C are the most important. The very size and variety of the P450 superfamily ensures that we do not need new enzymes for every existing or yet-to-be synthesised drug. Induction and inhibition of P450 enzymes is a fruitful source of drug–drug interactions.¹⁶

Each P450 enzyme protein is encoded by a separate gene (57 have been identified in humans), and variation in genes leads to differences between individuals, and sometimes between ethnic groups, in the ability to metabolise drugs. Persons who exhibit polymorphisms (see p. 107) inherit diminished or increased ability to metabolise substrate drugs, predisposing to toxicity or lack of efficacy.

Phase I oxidation of some drugs results in the formation of epoxides, which are short-lived and highly reactive metabolites that bind irreversibly through covalent bonds to cell constituents and are toxic to body tissues. Glutathione is a tripeptide that combines with epoxides, rendering them inactive, and its presence in the liver is part of an important defence mechanism against hepatic damage by halothane and paracetamol.

Note that some drug oxidation reactions do not involve the P450 system: several biologically active amines are inactivated by monoamine oxidase (see p. 319) and methylxanthines (see p. 154); mercaptopurine by xanthine oxidase (see p. 250); ethanol by alcohol dehydrogenase (see p. 143).

Hydrolysis (Phase I) reactions create active sites for subsequent conjugation of, e.g., aspirin, lidocaine, but this does not occur with all drugs.

Phase II

metabolism involves combination of the drug with one of several polar (water-soluble) endogenous molecules (products of intermediary metabolism), often at the active site (hydroxyl, amino, thiol) created by Phase I metabolism. The kidney readily eliminates the resulting water-soluble conjugate, or the bile if the molecular weight exceeds 300. Morphine, paracetamol and salicylates form conjugates with glucuronic acid (derived from glucose); oral contraceptive steroids form sulphates; isoniazid, phenelzine and dapsone are acetylated. Conjugation with a more polar molecule is also an elimination mechanism for natural substances, e.g. bilirubin as glucuronide, oestrogens as sulphates.

Phase II metabolism almost invariably terminates biological activity.

Transporters ¹⁷

It is convenient here to introduce the subject of carrier-mediated transporter processes whose physiological functions include the passage of amino acids, lipids, sugars, hormones and bile acids across cell membranes, and the protection of cells against environmental toxins.

There is an emerging understanding that membrane transporters have a key role in the overall disposition of drugs to their targeted organs. There are broadly two types: uptake transporters, which facilitate, for example, the passage of organic anions and cations into cells, and efflux transporters, which transport substances out of cells, often against high concentration gradients. Some transporters possess both influx and efflux properties.

Most efflux transporters are members of the ATP-binding cassette (ABC) superfamily that utilises energy derived from the hydrolysis of ATP; they include the P-glycoprotein family that expresses multidrug resistance protein 1 (MDR1) (see p. 516).

Their varied locations illustrate the potential for transporters widely to affect the distribution of drugs, namely in:

• Enterocytes of the small intestine, controlling absorption and thus bioavailability, e.g. of ciclosporin, digoxin.

• Liver cells, controlling uptake from the blood and excretion into the bile, e.g. of pravastatin.

• Renal tubular cells, controlling uptake from the blood, secretion into tubular fluid (and thus excretion) of organic anions, e.g. β-lactam antibiotics, diuretics, non-steroidal anti-inflammatory drugs.

• Brain capillary endothelial cells, controlling passage across the blood–brain barrier, e.g. of levodopa (but not dopamine) for benefit in Parkinson’s disease (see p. 361).

In time, it is likely that drug occupancy of transporter processes will provide explanations for some drug-induced toxicities and for a number of drug–drug interactions.

Enzyme induction

The mechanisms that the body evolved over millions of years to metabolise foreign substances now enable it to meet the modern environmental challenges of tobacco smoke, hydrocarbon pollutants, insecticides and drugs. At times of high exposure, our enzyme systems respond by increasing in amount and so in activity, i.e. they become induced; when exposure falls off, enzyme production gradually lessens.

A first alcoholic drink taken after a period of abstinence from alcohol may have a noticeable effect on behaviour, but the same drink taken at the end of 2 weeks of regular drinking may pass almost unnoticed because the individual’s liver enzyme activity is increased (induced), and alcohol is metabolised more rapidly, having less effect, i.e. tolerance is acquired. There is, nevertheless, a ceiling above which alcohol metabolising enzymes are not further induced.

Inducing substances

in general share some important properties: they tend to be lipid soluble, are substrates, though sometimes only minor ones, e.g. DDT, for the enzymes they induce, and generally have a long t_½. The time for onset and offset of induction depends on the rate of enzyme turnover, but significant induction generally occurs within a few days and it passes off over 2–3 weeks following withdrawal of the inducer.

Thus, certain drugs can alter the capacity of the body to metabolise other substances including drugs, especially in long-term use; this phenomenon has implications for drug therapy. More than 200 substances induce enzymes in animals but the list of proven enzyme inducers in humans is more restricted, as set out below.

Substances that cause enzyme induction in humans
• barbecued meats

• DDT (dicophane, and other insecticides)

• rifampicin

• ethanol (chronic use)

Enzyme induction is relevant to drug therapy because:

• Clinically important drug–drug (and drug–herb ¹⁸) interactions may result, for example, in failure of oral contraceptives, loss of anticoagulant control, failure of cytotoxic chemotherapy.

• Disease may result. Antiepilepsy drugs accelerate the breakdown of dietary and endogenously formed vitamin D, producing an inactive metabolite – in effect a vitamin D deficiency state, which can result in osteomalacia. The accompanying hypocalcaemia can increase the tendency to fits and a convulsion may lead to fracture of the demineralised bones.

• Tolerance to drug therapy may result in and provide an explanation for suboptimal treatment, e.g. with an antiepilepsy drug.

• Variability in response to drugs is increased. Enzyme induction caused by heavy alcohol drinking or heavy smoking may be an unrecognised cause for failure of an individual to achieve the expected response to a normal dose of a drug, e.g. warfarin, theophylline.

• Drug toxicity may occur. A patient who becomes enzyme induced by taking rifampicin is more likely to develop liver toxicity after paracetamol overdose by increased production of a hepatotoxic metabolite. (Such a patient will also present with a deceptively low plasma concentration of paracetamol due to accelerated metabolism; see p. 246).

Enzyme inhibition

The consequences of inhibiting drug metabolism can be more profound and more selective than enzyme induction because the outcome is prolongation of action of a drug or metabolite. Consequently, enzyme inhibition offers more scope for therapy (Table 8.4). Enzyme inhibition by drugs is also the basis of a number of clinically important drug interactions (see p. 107).

Table 8.4 Some drugs that act by enzyme inhibition

Drug	Enzyme inhibited	In treatment of
Acetazolamide	Carbonic anhydrase	Glaucoma
Allopurinol	Xanthine oxidase	Gout
Benserazide	DOPA decarboxylase	Parkinson’s disease
Disulfiram	Aldehyde dehydrogenase	Alcoholism
Enalapril	Angiotensin-converting enzyme	Hypertension, cardiac failure
Moclobemide	Monoamine oxidase, A type	Depression
Non-steroidal anti-inflammatory drugs	cyclo-oxygenase	Pain, inflammation
Selegiline	Monoamine oxidase, B type	Parkinson’s disease

Elimination

The body eliminates drugs following their partial or complete conversion to water-soluble metabolites or, in some cases, without their being metabolised. To avoid repetition the following account refers to the drug whereas the processes deal with both drug and metabolites.

Renal elimination

The following mechanisms are involved.

Glomerular filtration

The rate at which a drug enters the glomerular filtrate depends on the concentration of free drug in plasma water and on its molecular weight. Substances having a molecular weight in excess of 50 000 do not cross into the glomerular filtrate, whereas those of molecular weight less than 10 000 (which includes almost all drugs)¹⁹ pass easily through the pores of the glomerular membrane.

Renal tubular transport

Uptake and efflux transporters in proximal renal tubule cells transfer organic anions and cations between the plasma and the tubular fluid (see p. 453).

Renal tubular diffusion

The glomerular filtrate contains drug at the same concentration as it is free in the plasma, but the fluid is concentrated progressively as it flows down the nephron so that a gradient develops, drug in the tubular fluid becoming more concentrated than in the blood perfusing the nephron. As the tubular epithelium has the properties of a lipid membrane, the extent to which a drug diffuses back into the blood will depend on its lipid solubility, i.e. on its pK_a in the case of an electrolyte, and on the pH of tubular fluid. If the fluid becomes more alkaline, an acidic drug ionises, becomes less lipid soluble and its reabsorption diminishes, but a basic drug becomes un-ionised (and therefore more lipid soluble) and its reabsorption increases. Manipulation of urine pH gains useful expression with sodium bicarbonate given to alkalinise the urine for salicylate overdose.

Faecal elimination

When any drug intended for systemic effect is taken by mouth a proportion may remain in the bowel and be excreted in the faeces. Some drugs are intended not be absorbed from the gut, as an objective of therapy, e.g. neomycin. The cells of the intestinal epithelium contain several carrier-mediated transporters that control the absorption of drugs. The efflux transporter MDR1, for example, drives drug from the enterocyte into the gut lumen, limiting its bioavailability (see p. 93). Drug in the blood may also diffuse passively into the gut lumen, depending on its pK_a and the pH difference between blood and gut contents. The effectiveness of activated charcoal by mouth for drug overdose depends partly on its adsorption of such diffused drug, and subsequent elimination in the faeces (see p. 125).

Biliary excretion

Transporters regulate the uptake of organic cations and anions from portal blood to hepatocyte, and thence to the bile (see p. 86). The bile canaliculi tend to reabsorb small molecules and in general, only compounds having a molecular weight greater than 300 pass into bile. (See also Enterohepatic circulation, p. 86.)

Pulmonary elimination

The lungs are the main route of elimination (and of uptake) of volatile anaesthetics. Apart from this, they play only a trivial role in drug elimination. The route, however, acquires notable medicolegal significance when ethanol concentration is measured in the air expired by vehicle drivers involved in road traffic accidents (via the breathalyser).

Clearance

Elimination of a drug from the plasma is quantified in terms of its clearance. The term has the same meaning as the familiar renal creatinine clearance, which is a measure of removal of endogenous creatinine from the plasma. Clearance values can provide useful information about the biological fate of a drug. There are pharmacokinetic methods for calculating total body and renal clearance, and the difference between these represents hepatic clearance. The renal clearance of a drug eliminated only by filtration by the kidney obviously cannot exceed the glomerular filtration rate (adult male 124 mL/min, female 109 mL/min). If a drug has a renal clearance in excess of this, then the kidney tubules must actively secrete it, e.g. benzylpenicillin (renal clearance 480 mL/min).

Breast milk

Most drugs that are present in a mother’s plasma appear to some extent in her milk, although the amounts are so small that loss of drug in milk is of no significance as a mechanism of elimination. Even small amounts, however, may sometimes be of significance for the suckling child, whose drug metabolic and eliminating mechanisms are immature.

While most drugs taken by the mother pose no hazard to the child, exceptions to this observation occur because some drugs are inherently toxic, or transfer to milk in significant amounts, or there are known adverse effects, as below.

Drugs and breast feeding ²⁰

• Alimentary tract. Sulfasalazine may cause adverse effects and mesalazine appears preferable.

• Anti-asthma. The neonate eliminates theophylline and diprophylline slowly; observe the infant for irritability or disturbed sleep.

• Anticancer. Regard all as unsafe because of inherent toxicity.

• Antidepressants. Avoid doxepin, a metabolite of which may cause respiratory depression.

• Anti-arrhythmics (cardiac). Amiodarone is present in high and disopyramide in moderate amounts.

• Antiepilepsy. General note of caution: observe the infant for sedation and poor suckling. Primidone, ethosuximide and phenobarbital are present in milk in high amounts; phenytoin and sodium valproate less so.

• Anti-inflammatory. Regard aspirin (salicylates) as unsafe (possible association with Reye’s syndrome).

• Antimicrobials. Metronidazole is present in milk in moderate amounts; avoid prolonged exposure (though no harm recorded). Avoid nalidixic acid and nitrofurantoin where glucose-6-phosphate dehydrogenase deficiency is prevalent. Avoid clindamycin, dapsone, lincomycin, sulphonamides. Regard chloramphenicol as unsafe.

• Antipsychotics. Phenothiazines, butyrophenones and thioxanthenes are best avoided unless the indications are compelling: amounts in milk are small but animal studies suggest adverse effects on the developing nervous system. In particular, moderate amounts of sulpiride enter milk. Avoid lithium if possible.

• Anxiolytics and sedatives. Benzodiazepines are safe if use is brief but prolonged use may cause somnolence or poor suckling.

• β-Adrenoceptor blockers. Neonatal hypoglycaemia may occur. Sotalol and atenolol are present in the highest amounts in this group.

• Hormones. Oestrogens, progestogens and androgens suppress lactation in high dose. Oestrogen–progestogen oral contraceptives are present in amounts too small to be harmful, but may suppress lactation if it is not well established.

• Miscellaneous. Bromocriptine suppresses lactation. Caffeine may cause infant irritability in high doses.

Drug dosage

Drug dosage can be of five main kinds.

Fixed dose. The effect that is desired can be obtained at well below the toxic dose (many mydriatics, analgesics, oral contraceptives, antimicrobials) and enough drug can be given to render individual variation clinically insignificant.

Variable dose – with crude adjustments. Here fine adjustments make comparatively insignificant differences and the therapeutic endpoint may be hard to measure (depression, anxiety), may change only slowly (thyrotoxicosis), or may vary because of pathophysiological factors (analgesics, adrenal corticosteroids for suppressing disease).

Variable dose – with fine adjustments. Here a vital function (blood pressure, blood sugar level), which often changes rapidly in response to dose changes and can easily be measured repeatedly, provides the endpoint. Adjustment of dose must be accurate. Adrenocortical replacement therapy falls into this group, whereas adrenocortical pharmacotherapy falls into the group above.

Maximum tolerated dose is used when the ideal therapeutic effect cannot be achieved because of the occurrence of unwanted effects (anticancer drugs; some antimicrobials). The usual way of finding this is to increase the dose until unwanted effects begin to appear and then to reduce it slightly, or to monitor the plasma concentration.

Minimum tolerated dose. This concept is less common than the one above, but it applies to long-term adrenocortical steroid therapy against inflammatory or immunological conditions, e.g. in asthma and some cases of rheumatoid arthritis, when the dose that provides symptomatic relief may be so high that serious adverse effects are inevitable if it is continued indefinitely. The compromise is incomplete relief on the grounds of safety. This can be difficult to achieve.

Dosing schedules

Dosing schedules are simply schemes aimed at achieving a desired effect while avoiding toxicity. The following discussion assumes that drug effect relates closely to plasma concentration, which in turn relates closely to the amount of drug in the body. The objectives of a dosing regimen where continuing effect is required are:

To specify an initial dose

that attains the desired effect rapidly without causing toxicity. Often the dose that is capable of initiating drug effect is the same as that which maintains it. On repeated dosing, however, it takes 5 × t_½ periods to reach steady-state concentration in the plasma and this lapse of time may be undesirable. The effect may be achieved earlier by giving an initial dose that is larger than the maintenance dose; the initial dose is then called the priming or loading dose, i.e. the dose that will achieve a therapeutic effect in an individual whose body does not already contain the drug.

To specify a maintenance dose:

amount and frequency. Intuitively the maintenance dose might be half of the initial/priming dose at intervals equal to its plasma t_½, for this is the time by which the plasma concentration that achieves the desired effect, declines by half. Whether or not this approach is satisfactory or practicable, however, depends very much on the t_½ itself, as is illustrated by the following cases:

1. Half-life 6–12 h. In this instance, replacing half the initial dose at intervals equal to the t_½ can indeed be a satisfactory solution because dosing every 6–12 h is acceptable.

2. Half-life greater than 24 h. With once-daily dosing (which is desirable for compliance), giving half the priming dose every day means that more drug is entering the body than is leaving it each day, and the drug will accumulate to give unwanted effects. The solution is to replace only the amount of drug that leaves the body in 24 h, calculated from the inital dose, dose interval, and t_½.

3. Half-life less than 3 h. Dosing at intervals equal to the t_½ would be so frequent as to be unacceptable. The answer is to use continuous intravenous infusion if the t_½ is very short, e.g. dopamine t_½ 2 min (steady-state plasma concentration will be reached in 5 × t_½ = 10 min), or, if the t_½ is longer, e.g. lidocaine (t_½ 90 min), to use a priming dose as an intravenous bolus followed by a constant intravenous infusion. Intermittent administration of a drug with short t_½ is nevertheless reasonable provided large fluctuations in plasma concentration are acceptable, i.e. that the drug has a large therapeutic index. Benzylpenicillin has a t_½ of 30 min but is effective in a 6-hourly regimen because the drug is so non-toxic that it is possible safely to give a dose that achieves a plasma concentration many times in excess of the minimum inhibitory concentration for sensitive organisms.

Dose calculation by body-weight and surface area

A uniform, fixed drug dose is likely to be ineffective or toxic in several circumstances, e.g. cytotoxic chemotherapy, aminoglycoside antibiotics. It is usual then to calculate the dose according to body-weight. Adjustment according to body surface area is also used and may be more appropriate, for this correlates better with many physiological phenomena, e.g. metabolic rate.

The relationship between body surface area and weight is curvilinear, but a reasonable approximation is that a 70-kg human has a body surface area of 1.8 m². A combination of body-weight and height gives a more precise value for surface area (obtained from standard nomograms) and other more sophisticated methods.²¹

The issue takes on special significance for children, if the only dose known is that for the adult; adjustment is then commonly made by body-weight, or body surface area, among other factors (see p. 104).

Prolongation of drug action

Giving a larger dose is the most obvious way to prolong a drug action but this is not always feasible, and other mechanisms are used:

• Vasoconstriction will reduce local blood flow so that distribution of drug away from an injection site is retarded, e.g. combination with adrenaline/epinephrine prolongs local anaesthetic action.

• Slowing of metabolism may usefully extend drug action, as when a dopa decarboxylase inhibitor, e.g. carbidopa, is combined with levodopa (as co-careldopa) for parkinsonism.

• Delayed excretion is seldom practicable, the only important example being the use of probenecid to block renal tubular excretion of penicillin for single-dose treatment of gonorrhoea.

• Altered molecular structure can prolong effect, e.g. the various benzodiazepines.

• Pharmaceutical formulation. Manipulating the form in which a drug is presented by modified-release ²² systems can achieve the objective of an even as well as a prolonged effect.

Sustained-release (oral) preparations

can reduce the frequency of medication to once a day, and compliance becomes easier for the patient. The elderly can now receive most long-term medication as a single morning dose. In addition, sustained-release preparations may avoid bowel toxicity due to high local concentrations, e.g. ulceration of the small intestine with potassium chloride tablets; they may also avoid the toxic peak plasma concentrations that can occur when dissolution of the formulation, and so absorption of the drug, is rapid. Some sustained-release formulations also contain an immediate-release component to provide rapid, as well as sustained, effect.

Depot (injectable) preparations

are more reliable because the environment in which they are deposited is more constant than can ever be the case in the alimentary tract, and medication can be given at longer intervals, even weeks. In general, such preparations are pharmaceutical variants, e.g. micro-crystals, or the original drug in oil, wax, gelatin or synthetic media. They include phenothiazine neuroleptics, the various insulins and penicillins, preparations of vasopressin, and medroxyprogesterone (intramuscular, subcutaneous). Tablets of hormones can be implanted subcutaneously. The advantages of infrequent administration and better patient adherence in a variety of situations are obvious.

Reduction of absorption time

A soluble salt of the drug may be effective by being rapidly absorbed from the site of administration. In the case of subcutaneous or intramuscular injections, the same objective may be obtained with hyaluronidase, an enzyme that depolymerises hyaluronic acid, a constituent of connective tissue that prevents the spread of foreign substances, e.g. bacteria, drugs. Hyaluronidase combined with an intramuscular injection, e.g. a local anaesthetic, or a subcutaneous infusion leads to increased permeation with more rapid absorption. Hyaluronidase also promotes resorption of tissue accumulation of blood and fluid.

Fixed-dose drug combinations

This refers to combinations of drugs in a single pharmaceutical formulation. It does not mean concomitant drug therapy, e.g. in infections, hypertension and in cancer, when several drugs are given separately. Therapeutic aims should be clear. Combinations are logical if there is good reason to consider that the patient needs all the drugs in the formulation and that the doses are appropriate and will not need adjustment separately. Fixed-dose drug combinations are appropriate for:

• Convenience, with improved patient compliance, is appropriate with two drugs used at constant dose, long term, for an asymptomatic condition, e.g. a thiazide plus an ACE inhibitor in mild or moderate hypertension, and other antihypertensive drug combinations.

• Enhanced effect. Single-drug treatment of tuberculosis leads to the emergence of resistant mycobacteria and is prevented or delayed by using two or more drugs simultaneously. Combining isoniazid with rifampicin (Rifinah, Rimactazid) ensures that single-drug treatment cannot occur; treatment has to be two drugs or no drug at all. An oestrogen and progestogen combination provides effective oral contraception, for the same reason.

• Minimisation of unwanted effects. Levodopa combined with benserazide (Madopar) or with carbidopa (Sinemet) slows its metabolism outside the CNS so that smaller amounts of levodopa can be used, reducing its adverse effects.

Chronic pharmacology

The pharmacodynamics and pharmacokinetics of many drugs differ according to whether their use is in a single dose, or over a brief period (acute pharmacology), or long term (chronic pharmacology). An increasing proportion of the population take drugs continuously for large portions of their lives, as tolerable suppressive and prophylactic remedies for chronic or recurrent conditions are developed; e.g. for arterial hypertension, diabetes mellitus, mental diseases, epilepsies. In general, the dangers of a drug therapy are not markedly greater if therapy lasts for years rather than months, but long-term treatment can introduce significant hazard into patients’ lives unless management is skilful.

Interference with self-regulating systems

When self-regulating physiological systems (generally controlled by negative feedback systems, e.g. endocrine, cardiovascular) are subject to interference, their control mechanisms respond to minimise the effects of the interference and to restore the previous steady state or rhythm; this is homeostasis. The previous state may be a normal function, e.g. ovulation (a rare example of a positive feedback mechanism), or an abnormal function, e.g. high blood pressure. If the body successfully restores the previous steady state or rhythm then the subject has become tolerant to the drug, i.e. needs a higher dose to produce the desired previous effect.

In the case of hormonal contraceptives, persistence of suppression of ovulation occurs and is desired, but persistence of other effects, e.g. on blood coagulation and metabolism, is not desired.

In the case of arterial hypertension, tolerance to a single drug commonly occurs, e.g. reduction of peripheral resistance by a vasodilator is compensated by an increase in blood volume that restores the blood pressure; this is why a diuretic is commonly used together with a vasodilator in therapy.

Feedback systems

The endocrine system serves fluctuating body needs. Glands are therefore capable either of increasing or decreasing their output by means of negative (usually) feedback systems. An administered hormone or hormone analogue activates the receptors of the feedback system so that high doses cause suppression of natural production of the hormone. On withdrawal of the administered hormone, restoration of the normal control mechanism takes time, e.g. the hypothalamic–pituitary–adrenal cortex system can take months to recover full sensitivity, and sudden withdrawal of administered corticosteroid can result in an acute deficiency state that may be life endangering.

Regulation of receptors

The number (density) of receptors on cells (for hormones, autacoids or local hormones, and drugs), the number occupied (receptor occupancy) and the capacity of the receptor to respond (affinity, efficacy) can change in response to the concentration of the specific binding molecule or ligand,²³ whether this be agonist or antagonist (blocker). The effects always tend to restore cell function to its normal or usual state. Prolonged high concentrations of agonist (whether administered as a drug or over-produced in the body by a tumour) cause a reduction in the number of receptors available for activation (down-regulation); changes in receptor occupancy and affinity and the prolonged occupation of receptors antagonists lead to an increase in the number of receptors (up-regulation). At least some of this may be achieved by receptors moving inside the cell and out again (internalisation and externalisation).

Down-regulation, and the accompanying receptor changes, may explain the ‘on–off’ phenomenon in Parkinson’s disease (see p. 362) and the action of luteinising hormone releasing hormone (LHRH) super-agonists in reducing follicle stimulating hormone (FSH) concentrations for treating endocrine-sensitive prostate cancer.

Up-regulation. The occasional exacerbation of ischaemic cardiac disease on sudden withdrawal of a β-adrenoceptor blocker may be explained by up-regulation during its administration, so that, on withdrawal, an above-normal number of receptors suddenly become accessible to the normal transmitter, i.e. noradrenaline/norepinephrine.

Up-regulation with rebound sympathomimetic effects may be innocuous to a moderately healthy cardiovascular system, but the increased oxygen demand of these effects can have serious consequences where ischaemic disease is present and increased oxygen need cannot be met (angina pectoris, arrhythmia, myocardial infarction). Unmasking of a disease process that has worsened during prolonged suppressive use of the drug, i.e. resurgence, may also contribute to such exacerbations.

The rebound phenomenon

is plainly a potential hazard and the use of a β-adrenoceptor blocker in the presence of ischaemic heart disease would be safer if rebound is eliminated. β-Adrenoceptor blockers that are not pure antagonists but have some agonist (sympathomimetic ischaemic) activity, i.e. partial agonists, may prevent the generation of additional adrenoceptors (up-regulation). Indeed, there is evidence that rebound is less or is absent with pindolol, a partial agonist β-adrenoceptor blocker.

Sometimes a distinction is made between rebound (recurrence at intensified degree of the symptoms for which the drug was given) and withdrawal syndrome (appearance of new additional symptoms). The distinction is quantitative and does not imply different mechanisms.

Rebound and withdrawal phenomena occur erratically. In general, they are more likely with drugs having a short t_½ (abrupt drop in plasma concentration) and pure agonist or antagonist action. They are less likely to occur with drugs having a long t_½ and (probably) with those having a mixed agonist–antagonist (partial agonist) action on receptors.

Abrupt withdrawal

Clinically important consequences occur, and might occur for a variety of reasons, e.g. a patient interrupting drug therapy to undergo surgery. The following are examples:

• Cardiovascular system: β-adrenoceptor blockers, antihypertensives (especially clonidine).

• Nervous system: all depressants (hypnotics, sedatives, alcohol, opioids), antiepileptics, antiparkinsonian agents, tricyclic antidepressants.

• Endocrine system: adrenal corticosteroids.

• Immune inflammation: adrenal corticosteroids.

Resurgence

of chronic disease, which has progressed in severity although its consequences have been wholly or partly suppressed, i.e. a catching-up phenomenon, is a possible outcome of discontinuing effective therapy, e.g. levodopa in Parkinson’s disease. Corticosteroid withdrawal in autoimmune disease may cause both resurgence and rebound.

Drug discontinuation syndromes,

i.e. rebound, withdrawal and resurgence (defined above) are phenomena that are to be expected. The exact mechanisms may remain obscure but clinicians have no reason to be surprised when they occur, and in the case of rebound they may wish to use gradual withdrawal wherever drugs are used to modify complex self-adjusting systems, and to suppress (without cure) chronic diseases.

Other aspects of chronic drug use

Metabolic changes

over a long period may induce disease, e.g. thiazide diuretics (diabetes mellitus), adrenocortical hormones (osteoporosis), phenytoin (osteomalacia). Drugs may also enhance their own metabolism, and that of other drugs (enzyme induction).

Specific cell injury

or cell functional disorder occurs with individual drugs or drug classes, e.g. tardive dyskinesia (dopamine receptor blockers), retinal damage (chloroquine, phenothiazines), retroperitoneal fibrosis (methysergide), non-steroidal anti-inflammatory drugs (nephropathy). Cancer may occur, e.g. with oestrogens (endometrium) and with immunosuppressive (anticancer) drugs.

Drug holidays

The term means the deliberate interruption of long-term therapy in order to restore sensitivity (which has been lost) or to reduce the risk of toxicity. Plainly, the need for holidays is a substantial disadvantage for any drug. Patients sometimes initiate their own drug holidays (see Patient compliance, p. 21).

Dangers of intercurrent illness

are particularly notable with anticoagulants, adrenal corticosteroids and immunosuppressives.

Dangers of interactions with other drugs, herbs or food

see Index for individual drugs.

Conclusions

Drugs not only induce their known listed primary actions, but may:

• evoke compensatory responses in the complex interrelated physiological systems they affect, and these systems need time to recover on withdrawal of the drug (gradual withdrawal is sometimes mandatory and never harmful)

• induce metabolic changes that may be trivial in the short term, but serious if they persist for a long time

• produce localised effects in specially susceptible tissues and induce serious cell damage or malfunction

• increase susceptibility to intercurrent illness and to interaction with other drugs that may be taken for new indications.

That such consequences occur with prolonged drug use need evoke no surprise. But a knowledge of physiology, pathology and pharmacology, combined with awareness that the unexpected can occur, will allow patients who require long-term therapy to be managed safely, or at least with minimum risk of harm.

Individual or biological variation

Prescribing for special risk groups

That individuals respond differently to drugs, both from time to time and from other individuals, is a matter of everyday experience. Doctors need to accommodate for individual variation, as it may explain both adverse response to a drug and failure of therapy. Sometimes there are obvious physical characteristics such as age, ethnicity (genetics) or disease that warn the prescriber to adjust drug dose, but there are no external features that signify, e.g. pseudocholinesterase deficiency, which causes prolonged paralysis after suxamethonium. An understanding of the reasons for individual variation in response to drugs is relevant to all who prescribe. Both pharmacodynamic and pharmacokinetic effects are involved, and the issues fall in two general categories: inherited influences and environmental and host influences.

Pharmacogenomics

Munir Pirmohamed

We are grateful to Professor Munir Pirmohamed, NHS Chair of Pharmacogenetics at the University of Liverpool, UK, for providing the following account.

‘Variability is the law of life, and as no two faces are the same, so no two bodies are alike, and no two individuals react alike, and behave alike …’ (Sir William Osler,1849–1919).

Introduction

The response to drugs varies widely among patients – for example, it has been estimated that only about 30% of patients respond to antidepressants. Adverse drug reactions are also common, accounting for 6.5% of admissions to hospital. Part of this variability is due to patient-related factors (non-compliance, smoking, alcohol, co-morbidities) and poor prescribing. However, a significant proportion of the variability, which varies from drug to drug, is due to genetic factors. This area of study is termed pharmacogenomics.

Pharmacogenomics can be defined as the study of the genomic basis of why individuals vary in their response (efficacy and/or toxicity) to drugs. This term is used interchangeably with pharmacogenetics. The first example can be traced back to the time of Pythagoras who described the phenomenon ascribed to red cell haemolysis in some Mediterranean populations eating fava beans (favism), which we now know to be due to glucose-6-phosphate dehydrogenase deficiency, the most common enzyme deficiency in man.

Sources of variability

In general, variability in drug response can be due to pharmacokinetic and/or pharmacodynamic factors (Fig. 8.4). Variability in the expression of the cytochrome P450 enzymes, which are responsible for Phase I drug metabolism, has been the focus of most of the work in pharmacokinetics. Cytochrome P450 2D6 (CYP2D6), for example, is one of the most variable P450 enzymes in man, is absent in 8% of the UK population, and is responsible for the metabolism of 25% of drugs, including CNS (antidepressants and antipsychotics) and cardiovascular (β-blockers and anti-arrhythmics) drugs. Much less work has been done on pharmacodynamic factors causing variation in drug response, but as drugs can affect almost any protein in the body, almost every gene may have an effect on how individual drugs vary in their response. It is important to note, however, that for most drugs variability in response is due to a combination of both pharmacokinetic and pharmacodynamic factors, both of which can be affected by environmental or genetic factors. Specific examples are provided below.

Fig. 8.4 Variation in drug response can be due to genetically determined factors in pharmacokinetic and pharmacodynamic processes in the body. Some examples are given in each box.

Identifying genetic variation

Much of the work in pharmacogenomics has been based on a study of candidate genes, i.e. a study of genes known to be involved in the pharmacokinetics of the drug and its mechanism of action. Specifically the focus has been on single nucleotide polymorphisms (SNPs), which are base substitutions occurring with a frequency of at least 1% in the human population. With the advances in technologies, more recent work has focused on genome-wide association studies (GWAS), i.e. a study of all genes in the human genome without any prior knowledge of the pharmacokinetic and pharmacodynamic parameters of the drug. GWAS is also likely to be surpassed by next generation sequencing technologies in the near future, which will provide an evaluation of common and rare variants in the human genome.

Examples of pharmacogenomic variation

Drug efficacy

Cancer therapy

Cancer is essentially a genetic disease with approximately 30–80 mutations per cancer. These mutations within the cancer genome (the somatic genome) also change the responsiveness of the cancer to therapy. The best example is that of trastuzumab (Herceptin) in breast cancer; this drug improves disease-free and overall survival in patients with HER2 gene amplification or over-expression of the protein on breast cancer cells. This adverse prognostic factor occurs in 20% of newly diagnosed breast cancers. In patients with colon cancer, the proto-oncogene KRAS acts as a downstream signal transducer of epidermal growth factor receptor (EGFR), but can undergo activation independently of EGFR in the presence of mutations. Inhibition of EGFR, for example using inhibitory monoclonal antibodies such as panitumumab and cetuximab, is better in patients with the wild-type KRAS gene than in those with mutations.

Warfarin

Warfarin is a narrow therapeutic index drug where individual daily dose requirements vary by at least 40-fold. Inability to maintain an INR between 2 and 3 can predispose to either thrombosis (INR<2) or haemorrhage (INR>3). Warfarin is metabolised by various P450 enzymes, the most important being CYP2C9. Two variants in the CYP2C9 gene (termed CYP2C9*2 and CYP2C9*3) reduce the activity of the enzyme and overall rate of the metabolic turnover of warfarin. The mode of action of warfarin is through interruption of the vitamin K cycle specifically by inhibiting the enzyme vitamin K epoxide reductase complex 1 (VKORC1) – variation in this gene can affect the daily requirements for warfarin. In most global populations, it has now been shown that age and body mass index, together with genetic variation in CYP2C9 and VKORC1, can account for at least 50% of the variation in daily dose requirements for warfarin. This has resulted in the development of dosing algorithms in an attempt to improve the accuracy and prediction of individual dose requirements for warfarin.

Drug toxicity

Immune-mediated adverse drug reactions

Many type B or idiosyncratic adverse drug reactions are immunologically mediated. Immune response to antigens, including those derived from drugs, is partly under the control of the HLA genes on chromosome 6, which is the polymorphic region of the human genome. Not surprisingly, HLA genes are now being found to be important determinants of susceptibility to these immune-mediated adverse reactions. The best example of this is with abacavir, a drug used to treat HIV, which causes hypersensitivity (skin rash, fever, gastrointestinal and respiratory manifestations) in 5% of patients. A strong association of abacavir hypersensitivity with the HLA allele, HLA-B*5701, has now been shown in several populations. Furthermore, genotyping for HLA-B*5701 before prescribing abacavir has been shown to reduce the frequency of hypersensitivity, and is a cost-effective approach. In Europe, it is now mandatory to undertake HLA-B*5701 testing before the prescription of abacavir. Very strong genetic associations between the HLA genes and different forms of hypersensitivity, including those affecting the skin and liver, occur with a number of drugs (Table 8.5).

Table 8.5 HLA alleles predisposing to immune-mediated adverse drug reactions

Statin myopathy

Statins are among the most widely used drugs in the world, with marked benefits in terms of reduction of cardiovascular morbidity and mortality. Although generally well tolerated, statins can occasionally cause muscle damage, which can range from muscle pains to rhabdomyolysis (associated with renal failure which can be fatal). A genome-wide approach in patients on simvastatin showed that a variant in an influx drug transporter called SLCO1B1 (also known as OATP1B1) increased the risk of statin-related myopathy. This transporter is highly expressed in the liver and is responsible for the transport of some statins from the blood into the hepatocyte. In patients expressing this variant, the activity of the transporter is reduced which leads to a decrease in hepatocyte uptake and an increase in plasma concentrations of the statin. However, the mechanism by which the statin leads to muscle damage is unknown.

Summary

There are many genetic variations identified to be risk factors for lack of efficacy or predisposition to toxicity (Table 8.6). As the technologies to interrogate the human genome improve, it is likely that more genetic tests will be introduced which will need to be used before the prescription of the drug. The net effect will be prediction of individual responses and thereby reduction in variability through better drug choices and/or drug doses.

Table 8.6 Drugs which contain pharmacogenetic information in their product labels

Drug	Drug class	Genomic variation
Maraviroc	Antiretroviral, antagonist of the CC chemokine receptor 5 (CCR5)	CCR5 promoter and coding sequence polymorphisms
Trastuzumab (Herceptin)	Anticancer drug, anti-HER-2/neu monoclonal antibody used where there is over-expression of the human epidermal growth factor receptor-2 (HER2)	HER2/neu
Abacavir	Antiretroviral, nucleoside reverse transcriptase inhibitor	Human leucocyte antigen HLA-B*5701 allele
Carbamazepine	Antiepileptic	HLA-B*1502 in patients of Asian ancestry
Warfarin	Anticoagulans, vitamin K epoxide reductase inhibitor	CYP 2 C92 and 2 C93 and VKORC1 variants
Azathioprine	Antiproliferative immunosuppressant	Thiopurine methyltransferase (TPMT) deficiency
Valproic acid	Antiepileptic and antimanic drug	Urea cycle disorder (UCD) deficiency
Hydralazine	Vasodilator antihypertensive drug	N-acetyl transferase (NAT)
Rifampicin Isoniazid	Antituberculous drug	N-acetyl transferase (NAT)
Voriconazole	Antifungal	CYP 2 C19 variants poor and extensive metabolisers
Diazepam	Anxiolytic	CYP 2 C19 variants poor and extensive metabolisers
Fluoxetine	Selective serotonin reuptake inhibitor	Cytochrome P450 CYP 2D6, substrate and inhibitor
Tramadol	Analgesic	CYP 2D6
Propranolol	β-Adrenoceptor blocking drug	CYP 2D6
Tamoxifen	Oestrogen-receptor antagonist	CYP 2D6
Tretinoin	Acid form of vitamin A used in acute promyelocytic leukaemia	Presence of the t(15;17) translocation and/or PML/RARα gene fusion
Celecoxib	Non-steroidal anti-inflamatory drug, selective COX-2 inhibitor	CYP 2 C9 variants with poor metaboliser status
Primaquine Chloroquine	Antimalarials	Glucose-6-phosphate dehydrogenase (G6PD) deficiency
Suxamethonium	Anaesthetics	Butyrylcholinesterase deficiency

Environmental and host influences

A multitude of factors related to both individuals and their environment contribute to differences in drug response. Some of the more relevant influences are the following:

Age

The neonate, infant and child ²⁴

Young human beings differ greatly from adults, not merely in size but also in the proportions and constituents of their bodies and the functioning of their physiological systems. These differences influence the way the body handles and responds to drugs:

• Rectal absorption is efficient with an appropriate formulation, e.g. of diazepam and theophyllines; this route may be preferred with an uncooperative infant.

• The intramuscular or subcutaneous routes tend to give unpredictable plasma concentrations, e.g. of digoxin or gentamicin, because of the relatively low proportion of skeletal muscle and fat. Intravenous administration is preferred in the seriously ill newborn.

• Drugs or other substances that come in contact with the skin are readily absorbed as the skin is well hydrated and the stratum corneum is thin; overdose toxicity may result, e.g. with hexachlorophene used in dusting powders and emulsions to prevent infection.

An understandable reluctance to test drugs extensively in children means that reliable information is often lacking. Many drugs do not have a licence to be used for children, and their prescription must be ‘off-licence’, a practice that is recognised as necessary, if not actually promoted, by the UK drug regulatory authorities.²⁵ Attempts to correct this are underway across Europe.

Distribution

Total body water in the neonate amounts to 80%, compared with 65% of body-weight in older children. Consequently:

• Less extensive binding of drugs to plasma proteins is generally without clinical importance but there is a risk of kernicterus in the jaundiced neonate following displacement of bilirubin from protein-binding sites by vitamin K, X-ray contrast media or indometacin.

Metabolism

Drug-inactivating enzyme systems are present at birth but are functionally immature (particularly in the preterm baby), especially for oxidation and for conjugation with glucuronic acid. Inadequate conjugation and thus inactivation of chloramphenicol by neonates causes the fatal ‘grey’ syndrome, but this is not a widely used antibiotic. After the initial weeks of life, because their drug metabolic capacity increases rapidly, young children may require a higher weight-related dose than adults.

Elimination

Glomerular filtration, tubular secretion and reabsorption are low in the neonate (even lower in preterm babies), reaching adult values in relation to body surface area only at 2–5 months. Drugs that the kidney excretes, e.g. aminoglycosides, penicillins, diuretics, require reduced dose; after about 6 months, body-weight- or surface area-related daily doses are the same for all ages.

Pharmacodynamic responses

There is scant information about developmental effects of interaction between drugs and receptors. Other sources suggest possible effects: e.g. thalidomide causes phocomelia only in the forming limb (see Index); tetracyclines stain only developing enamel; young children are particularly sensitive to liver toxicity from valproate.

Dosage in the young

No single rule or formula is suitable for all cases. Computation by body-weight may overdose an obese child, for whom calculation of ideal weight from age and height is preferred. Doses based on body surface area are generally more accurate, and preferably should take into account both body-weight and height.²⁶ The fact that the surface area of a 70-kg adult human is 1.8 m² (see p. 97) then allows adjustment, as follows:

General guidance is available from formularies, e.g. the British National Formulary, and specialist publications.²⁷^,²⁸

The elderly

The incidence of adverse drug reactions rises with age in the adult, especially after 65 years, because of:

• The increasing number of drugs that the elderly need because they tend to have multiple diseases.

• Poor compliance with dosing regimens.

• Bodily changes of ageing that require modified dosage regimens.

Absorption

of drugs administered orally may be slightly slower because of reduced gastrointestinal blood flow and motility but the effect is rarely important.

Distribution

reflects the following changes:

• Lean body mass is less and standard adult doses provide a greater amount of drug per kilogram.

• Body fat increases and may act as a reservoir for lipid-soluble drugs.

• Total body water is less and, in general, water-soluble drugs have a lower distribution volume. Standard doses of drugs, especially the loading doses of those that are water soluble, may thus exceed the requirement.

• Plasma albumin concentration is well maintained in the healthy elderly but may fall with chronic disease, giving scope for a greater proportion of unbound (free) drug, which may be important when priming doses are given.

Metabolism

reduces as liver mass and liver blood flow decline. Consequently:

• Metabolic inactivation of drugs is slower, mostly for Phase I (oxidation) reactions; the capacity for Phase II (conjugation) is better preserved.

• Drugs normally extensively eliminated in first pass through the liver appear in higher concentration in the systemic circulation and persist in it for longer. There is, therefore, particular need initially to use lower doses of most neuroleptics, tricyclic antidepressants and cardiac anti-arrhythmic agents.

• Capacity for hepatic enzyme induction appears less.

Elimination

Renal blood flow, glomerular filtration and tubular secretion decrease with age above 55 years, a decline that raised serum creatinine concentration does not signal because production of this metabolite is diminished by the age-associated diminution of muscle mass. Indeed, in the elderly, serum creatinine may be within the concentration range for normal young adults even when the creatinine clearance is 50 mL/min (compared with 127 mL/min in adult males). Particular risk of adverse effects arises with drugs that are eliminated mainly by the kidney and that have a small therapeutic ratio, e.g. aminoglycosides, digoxin, lithium.

Pharmacodynamic

response may alter with age, to produce either a greater or a lesser effect than is anticipated in younger adults, for example:

• Drugs that act on the CNS appear to produce an exaggerated response in relation to that expected from the plasma concentration, and sedatives and hypnotics may have a pronounced hangover effect. These drugs are also more likely to depress respiration because of reduced vital capacity and maximum breathing capacity in the elderly.

• Response to β-adrenoceptor agonists and antagonists may diminish in old age, possibly through reduced affinity for adrenoceptors, or smaller number of receptors.

• Baroreceptor sensitivity reduces, leading to greater potential for orthostatic hypotension with drugs that reduce blood pressure.

These pharmacokinetic and pharmacodynamic differences, together with broader issues particular to the elderly, influence the choice and use of drugs for this age group, as follows:

Rules of prescribing for the elderly ²⁹

1. Think about the necessity for drugs. Is the diagnosis correct and complete? Is the drug really necessary? Is there a better alternative?

2. Do not prescribe drugs that are not useful. Think carefully before giving an old person a drug that may have major side-effects, and consider alternatives, including prescribing nothing.

3. Think about the dose. Is it appropriate to possible alterations in the patient’s physiological state? Is it appropriate to the patient’s renal and hepatic function at the time?

4. Think about drug formulation. Is a tablet the most appropriate form of drug or would an injection, a suppository or a syrup be better? Is the drug suitably packaged for the elderly patient, bearing in mind any disabilities?

5. Assume any new symptoms may be due to drug side-effects or, more rarely, to drug withdrawal, unless shown to be otherwise. Rarely (if ever) treat a side-effect of one drug with another.

6. Take a careful drug history. Bear in mind the possibility of interaction with substances the patient may be taking without your knowledge, such as herbal or other non-prescribed remedies, old drugs taken from the medicine cabinet or drugs obtained from friends.

7. Use fixed combinations of drugs only when they are logical and well studied, and they either aid compliance or improve tolerance or efficacy. Few fixed combinations meet this standard.

8. When adding a new drug to the therapeutic regimen, see whether another can be withdrawn.

9. Attempt to check whether the patient’s compliance is adequate, e.g. by counting remaining tablets. Has the patient (or relatives) been properly instructed?

10. Remember that stopping a drug is as important as starting it.

The old (80 + years) are particularly intolerant of neuroleptics (given for confusion) and of diuretics (given for ankle swelling that is postural and not due to heart failure), which cause adverse electrolyte changes. Both classes of drug may result in admission to hospital of semi-comatose ‘senior citizens’ who deserve better treatment from their juniors.

Pregnancy

As pregnancy evolves, profound changes occur in physiology, including fluid and tissue composition.

Absorption

Despite reduced gastrointestinal motility, there appears to be no major defect in drug absorption except that slow gastric emptying delays the appearance in the plasma of orally administered drugs, especially during labour. Absorption from an intramuscular site is likely to be efficient because vasodilatation increases tissue perfusion.

Distribution

Total body water increases by up to 8 L, creating a larger space within which water-soluble drugs may distribute. Plasma albumin (normal 33–55 g/L) declines by some 10 g/L from haemodilution. While this gives scope for increased free concentration of drugs that normally bind to albumin, unbound drug is also available to distribute, be metabolised and excreted. With phenytoin, for example, the free (and pharmacologically active) concentration does not alter, despite the dilutional fall in the total plasma concentration.

Thus therapeutic drug monitoring interpreted by concentrations appropriate for non-pregnant women may mislead. A useful general guide during pregnancy is to maintain concentrations at the lower end of the recommended range. Body fat increases by about 4 kg and provides a reservoir for lipid-soluble drugs.

Hepatic metabolism

increases, although not blood flow to the liver. There is increased clearance of drugs such as phenytoin and theophylline, whose elimination depends on liver enzyme activity. Drugs that are so rapidly metabolised that elimination depends on delivery to the liver, i.e. on hepatic blood flow, have unaltered clearance, e.g. pethidine.

Elimination

Renal plasma flow almost doubles and there is more rapid loss of renally excreted drugs, e.g. amoxicillin, the dose of which should be doubled for systemic infections (but not for urinary tract infections as penicillins are highly concentrated in the urine).

Resection and reconstruction of the gut may lead to malabsorption, e.g. of iron, folic acid and fat-soluble vitamins after partial gastrectomy, and of vitamin B₁₂ after ileal resection. Delayed gastric emptying and intestinal stasis during an attack of migraine interfere with drug absorption. Severe low-output cardiac failure or shock (with peripheral vasoconstriction) delays absorption from subcutaneous or intramuscular sites; reduced hepatic blood flow prolongs the presence in the plasma of drugs that are so rapidly extracted by the liver that removal depends on their rate of presentation to it, e.g. lidocaine.

Distribution

Hypoalbuminaemia from any cause, e.g. burns, malnutrition, sepsis, allows a higher proportion of free (unbound) drug in plasma. Although free drug is available for metabolism and excretion, there remains a risk of enhanced or adverse responses especially with initial doses of those that are highly protein bound, e.g. phenytoin.

Metabolism

Acute inflammatory disease of the liver (viral, alcoholic) and cirrhosis affect both the functioning of the hepatocytes and blood flow through the liver. Reduced extraction from the plasma of drugs that are normally highly cleared in first pass through the liver results in increased systemic availability of drugs such as metoprolol, labetalol and clomethiazole. Many other drugs exhibit prolonged t_½ and reduced clearance in patients with chronic liver disease, e.g. diazepam, tolbutamide, rifampicin (see p. 87). Thyroid disease has the expected effects, i.e. drug metabolism accelerates in hyperthyroidism and decelerates in hypothyroidism.

Elimination

Renal disease has profound effects on the elimination and hence duration of action of drugs eliminated by the kidney (see p. 462).

Pharmacodynamic changes

• Asthmatic attacks can be precipitated by β-adrenoceptor blockers.

• Malfunctioning of the respiratory centre (raised intracranial pressure, severe pulmonary insufficiency) causes patients to be intolerant of opioids, and indeed any sedative may precipitate respiratory failure.

• Myocardial infarction predisposes to cardiac arrhythmia with digitalis glycosides or sympathomimetics.

• Myasthenia gravis is aggravated by quinine and quinidine, and myasthenics are intolerant of competitive neuromuscular blocking agents and aminoglycoside antibiotics.

Food

• The presence of food in the stomach, especially if it is fatty, delays gastric emptying and the absorption of certain drugs, e.g. ampicillin and rifampicin. More specifically, calcium, for instance in milk, interferes with absorption of tetracyclines and iron (by chelation).

• Substituting protein for fat or carbohydrate in the diet is associated with an increase in drug oxidation rates. Some specific dietary factors induce drug metabolising enzymes, e.g. alcohol, charcoal grilled (broiled) beef, cabbage and Brussels sprouts.

Protein malnutrition causes changes that are likely to influence pharmacokinetics, e.g. loss of body-weight, reduced hepatic metabolising capacity, hypoproteinaemia.

Citrus flavinoids in grapefruit (but not orange) juice decrease hepatic metabolism and may lead to toxicity from amiodarone, terfenadine (cardiac arrhythmia), benzodiazepines (increased sedation), ciclosporin, felodipine (reduced blood pressure).

Drug interactions

When a drug is administered, a response occurs; if a second drug is given and the response to the first drug is altered, a drug–drug interaction is said to have occurred.

Dramatic unintended interactions excite most notice but they should not distract attention from the many intended interactions that are the basis of rational polypharmacy, e.g. multi-drug treatment of tuberculosis, naloxone for morphine overdose.

For completeness, alterations in drug action caused by diet (above) are termed drug–food interactions, and those by herbs drug–herb interactions.³⁰

Clinical importance of drug interactions

The quantity of drugs listed in any national formulary provides ample scope for possible alteration in the disposition or effect of one drug by another drug. But, in practice, clinically important adverse drug–drug interactions become likely with:

• Drugs that have a steep dose–response curve and a small therapeutic index (see p. 77) because small quantitative changes at the target site, e.g. receptor or enzyme, lead to substantial changes in effect, e.g. digoxin or lithium.

• Drugs that are known enzyme inducers or inhibitors (see pp. 93–94).

• Drugs that exhibit saturable metabolism (zero-order kinetics), when small interference with kinetics may lead to large alteration of plasma concentration, e.g. phenytoin, theophylline.

• Drugs that are used long term, where precise plasma concentrations are required, e.g. oral contraceptives, antiepilepsy drugs, cardiac anti-arrhythmia drugs, lithium.

• Severely ill patients, for they may be receiving several drugs; signs of iatrogenic disease may be difficult to distinguish from those of existing disease and the patient’s condition may be such that they cannot tolerate further adversity.

• Patients who have significantly impaired liver or kidney function, for these are the principal organs that terminate drug action.

• The elderly, for they tend to have multiple pathology, and may receive several drugs concurrently (see p. 105).

Pharmacological basis of drug interactions

Listings of recognised or possible adverse drug–drug interactions are now readily available in national formularies, on compact disk or as part of standard prescribing software. We provide here an overview of the pharmacological basis for wanted and unwanted, expected and unexpected effects when drug combinations are used.

Drug interactions are of two principal kinds:

• Pharmacodynamic interaction: both drugs act on the target site of clinical effect, exerting synergism (below) or antagonism. The drugs may act on the same or different receptors or processes, mediating similar biological consequences. Examples include: alcohol + benzodiazepine (to produce enhanced sedation), atropine + β-adrenoceptor blocker (to indirectly reverse β-adrenoceptor blocker overdose).

• Pharmacokinetic interaction: the drugs interact remotely from the target site to alter plasma (and other tissue) concentrations so that the amount of the drug at the target site of clinical effect is altered, e.g. enzyme induction by rifampicin reduces the plasma concentration of warfarin; enzyme inhibition by ciprofloxacin increases the concentration of theophylline.

Interaction may result in antagonism or synergism.

Antagonism

occurs when the action of one drug opposes that of another. The two drugs simply have opposite pharmacodynamic effects, e.g. histamine and adrenaline/epinephrine on the bronchi exhibit physiological or functional antagonism; or they compete reversibly for the same drug receptor, e.g. flumazenil and benzodiazepines exhibit competitive antagonism.

Synergism ³¹

is of two sorts:

1. Summation or addition occurs when the effects of two drugs having the same action are additive, i.e. 2 + 2 = 4 (a β-adrenoceptor blocker plus a thiazide diuretic have an additive antihypertensive effect).

2. Potentiation (to make more powerful) occurs when one drug increases the action of another, i.e. 2 + 2 = 5. Sometimes the two drugs both have the action concerned (trimethoprim plus sulfonamide), and sometimes one drug lacks the action concerned (benserazide plus levodopa), i.e. 0 + 2 = 3.

Strictly, the term synergism applies only to the second condition, but it is now commonly applied to both.

In broad terms, it is useful to distinguish the drug–drug interactions that occur:

• before drugs enter the body

• at important points during their disposition and metabolism

• at receptor sites.

Before administration

Intravenous fluids offer special scope for interactions (incompatibilities). Drugs commonly are weak organic acids or bases, produced as salts to improve their solubility. Plainly, the mixing of solutions of salts can result in instability, which may or may not be visible in the solution, i.e. precipitation. While specific sources of information are available in manufacturers’ package inserts and formularies, issues of compatibility are complex and lie within the professional competence of the hospital pharmacy, which should prepare drug additions to infused solutions. In any situation involving unfamiliar drugs their help and advice should be sought.

At the site of absorption

The complex environment of the gut provides opportunity for drugs to interfere with one another, both directly and indirectly, by altering gut physiology. Usually the result is to impair absorption.

By direct chemical interaction in the gut. Antacids that contain aluminium and magnesium form insoluble and non-absorbable complexes with tetracyclines, iron and prednisolone. Milk contains sufficient calcium to warrant its avoidance as a major article of diet with tetracyclines. Colestyramine interferes with the absorption of levothyroxine, digoxin and some acidic drugs, e.g. warfarin. Separating the dosing of interacting drugs by at least 2 h should largely avoid the problem.

By altering gut motility. Slowing of gastric emptying, e.g. opioid analgesics, tricyclic antidepressants (antimuscarinic effect), may delay and reduce the absorption of other drugs. Purgatives reduce the time spent in the small intestine and give less opportunity for the absorption of poorly soluble substances such as adrenal corticosteroids and digoxin.

By altering gut flora. Antimicrobials potentiate oral anticoagulants by reducing bacterial synthesis of vitamin K (usually only after antimicrobials are given orally in high dose, e.g. to treat Helicobacter pylori).

Interactions other than in the gut. Hyaluronidase promotes dissipation of a subcutaneous injection, and vasoconstrictors, e.g. adrenaline/epinephrine, felypressin, delay absorption of local anaesthetics, usefully to prolong local anaesthesia.

During distribution

Carrier-mediated transporters control processes such as bioavailability, passage into the CNS, hepatic uptake and entry into bile, and renal tubular excretion (see Index). Inhibitors and inducers of drug transporters can profoundly influence the disposition of drugs. The transporter MDR1 controls the entry of digoxin into cells; quinidine, verapamil and ciclosporin inhibit this transporter and increase the plasma concentration of digoxin (with potentially toxic effects). Probenecid inhibits the organic anion renal transporter, which decreases the renal clearance of penicillin (usefully prolonging its effect) but also that of methotrexate (with danger of toxicity). Elucidation of the location and function of transport systems will give the explanation for, and allow the prediction of, many more drug–drug interactions.