Principles of pharmacology and autonomic physiology relevant to anaesthesia and intensive care

Introduction

Only 25 years ago, the amount of information deemed necessary for efficient clinical use of a common drug would fit into a few lines of text. Approximately 500 new drugs have since been introduced for clinical use, and many more will be introduced in the future. Clinicians are now presented with much information on newly released drugs. This information is usually presented under three main headings.

Pharmaceutical

• Patient compliance, especially the elderly with multiple drug regimes
• How the drugs are presented and administered e.g. tablet, capsule, slow release and by what route
• Speed and extent of drug dispersal in the gastro-intestinal tract (GIT)
• Effects of other co-administered drugs and substances e.g. l dopa and carbidopa, tetracycline and milk

This is not of great relevance to anaesthesia but changes in the formulation of drugs (e.g. digoxin and phenytoin) have had profound effects on the absorption (and toxicity) of these drugs when given orally. Other factors include the storage and cost of drugs, the latter an extremely pertinent factor. It will not be considered further.

Pharmacokinetics

'What the body does to the drug'

i.e. how the body handles the drug following its administration until it is eliminated. This concerns absorption (e.g. from GIT, subcutaneous tissue, rectum), distribution, binding, metabolism and elimination. These factors can be quantified to give the pharmacokinetic parameters of half lives of absorption (T1/2 abs), redistribution (T1/2 alpha ) and elimination (T1/2 beta ), loading dose (Ld), maintenance dose (Md), clearance (Cl) and volume of distribution (Vd). These are explained more fully in the sections that follow.

Pharmacodynamics

'What the drug does to the body'

This describes the effect of the drug and how it works. It concerns theory of drug action, e.g. in terms of receptor occupancy, description of effects in normal and diseased states, plasma concentration - response relationships (i.e. the concept of effective concentration, Ceff), interference with other drugs and effects of overdose, interactions and adverse effects. Present standards of practice demand extensive knowledge of pharmacological principles. Anaesthesia provides excellent examples of the application of such principles.

Pharmacokinetic principles

All drugs need to cross at least one cell membrane to produce their desired effects. Factors affecting this process are:

These determine passage of drugs through pores in cell membranes, epithelia and endothelia. For example, early confinement of plasma expanders to the vascular compartment depends on molecular size. Molecular configuration of drugs determine binding to specific cell surface proteins for active transport across the membrane. Active transport is highly selective, saturable, requires energy (ATP Þ ADP + P), and may operate against a concentration gradient (e.g. amino acids, iodine). If it does not operate against a gradient it is called facilitated diffusion (e.g. glucose).

Lipid solubility is the most important determinant of diffusion across cell membranes because of the predominantly lipid nature of their constituents. Lipid-soluble drugs easily cross endothelial (e.g. GIT, blood-brain and placental) barriers according to the concentration gradient, the drugs dissolving initially in the lipid bilayer. The more lipid soluble the drug, the faster the rate of diffusion. However, cell membranes also contain protein molecules which float on this 'sea of lipid' and thus can also form hydrophilic channels to allow transport of different classes of small (MW < 100), water soluble drugs alongside the bulk flow of water which occurs across all membranes.

Most drugs are weak acids or bases and therefore dissociate into their unionised and ionised moieties in aqueous solution. Only the unionised fraction is lipophilic and easily diffusible across cell membranes. Acidic drugs are maximally unionised (lipophilic) at acid (low) pH, and basic drugs at alkaline (high) pH. The pKa of the drug is also important as it indicates the pH at which the drug is 50% unionised. Most acidic drugs have a pKa in the acid range (< 7.4) and vice versa for basic drugs. Thus, the proportion of a drug in the unionised, lipophilic form depends on the pKa of the drug, whether it is an acid or a base and the pH of the tissue or medium where it is present.

By rearranging the Henderson-Hasselbalch equation we get:

• e.g. aspirin has a pKa of 3.5: thus, if the pH of the stomach contents is 1.5, the log is 2. Thus, 100 times as much aspirin is present in the unionised as the ionised form in the stomach, absorption is enhanced. It also explains why alkalisation of the urine has the opposite effect in aspirin overdose, rendering the aspirin ionised so that it is less easily re absorbed by the kidney.

• e.g. pancuronium, being largely ionised at plasma pH, does not cross the blood-brain and placental barriers.

The degree of lipophilicity of the unionised fraction of a lipophilic drugs is quantified by it's ability to distribute in vitro between an octanol (oil) and water phase. The higher the octanol/water coefficient the more lipid soluble the drug and hence the easier it's passage across membranes.

Drugs may be given by a variety of routes. There is always some systemic absorption, even when drugs are applied topically.

Lipid solubility and ionisation are important for absorption (see above). Drugs given orally are mainly absorbed in the small intestine and then into the portal system, passing through the liver before reaching the general circulation. For lipophilic drugs which undergo significant hepatic biotransformation, this 'first pass' through the liver causes a large reduction in the amount of drug available (e.g. propranolol and most opioids). Sublingual (and to a much lesser effect, rectal) application bypass this effect. Bioavailability is the fraction of an oral dose reaching the systemic circulation as compared with the same amount given i.v.; it is affected by the combined effects of absorption and first pass metabolism. Paradoxically, a highly lipophilic drug may be completely absorbed from the small intestine but then undergo extensive first pass metabolism in the liver, thus reducing bioavailability.

Intravenous injection produces a peak plasma concentration within one circulation time (approximately one minute). In terms of effect, it is the most predictable route and the one usually used in emergencies. Drugs given by intramuscular and subcutaneous injection are variably absorbed depending on local tissue perfusion. The latter is markedly reduced in shocked states and this greatly delays onset.

Mucous membranes and skin

These are used as a route for systemic absorption as well as for local effects limited to the site of application:

Systemic use:

sublingual: buprenorphine, nifedipine and glyceryl trinitrate (GTN), thus avoiding hepatic first pass metabolism

rectum (as suppositories): the NSAID diclofenac is often given by this route but there is no evidence that bioavailability is any better than by the oral route.

skin: GTN, hyoscine, nicotine patches and fentanyl can be given by this route. It provides the equivalent of long term 'infusion' of drug, thus maintaining steady levels over many hours or days. It is only suitable for potent drugs as the amount of absorption through the skin is limited. In addition, absorption of active drug from skin depots continues for many hours following removal of the patch. This is obviously important with potent respiratory depressants such as fentanyl.

nasal mucosa: e.g. arginine vasopressin (DDAVP for diabetes inspidus) and fentanyl

trachea and large bronchi: is a particularly useful route for administration of adrenaline during cardiac arrest prior to obtaining venous access. Roughly twice the equivalent i.v. dose must be given by this route, but the amount absorbed is often unpredictable

inhalation: of gases and vapours are of particular interest in this context, e.g. the administration of inhalational anaesthetics.

Local use:

skin: A particularly good example in anaesthesia is the eutectic mixture of local anaesthetics (EMLA ), prilocaine and lignocaine. A eutectic mixture is a mixture of two compounds in a certain proportion which has the lowest melting point. For prilocaine and lignocaine in a 50:50 mixture, the melting point is lowered to below room temperature (individually 30 - 50 deg C), thus allowing the development of a very high concentration local anaesthetic cream which is capable of penetrating the skin.

Sometimes the systemic absorption of topically applied drugs (for local effects) can inadvertently produce systemic toxicity. A few examples are:

eye: plasma pseudocholinesterase may be inhibited by ecothiopate eye drops thus prolonging the effects of suxamethonium.

lung: salbutamol and other bronchodilators may be given in sufficient doses to produce toxicity (e.g. tachycardia and dysrhythmias)

Distribution and binding

Drugs are not evenly distributed throughout all tissues of the body. Well-perfused organs such as the brain and heart may initially be exposed to higher concentrations of drug before redistribution to other tissues.

A proportion may bind to plasma proteins; to albumin for acidic drugs and to alpha 1 - acid glycoprotein for basic drugs. Since only free, unbound drug is available for activity (or excretion by the kidney), the extent of protein binding influences the action of the drug. Some drugs compete for similar sites and can therefore displace each other leading to an enhanced effect of the displaced drug. The effect of this is slight in most instances as the free drug is then either rapidly redistributed or made available for metabolism and elimination.

The mathematics of pharmacokinetics is concerned with time-dependent changes in plasma concentration following administration of drugs (thus described as 'what the patient does to the drug'). Knowledge of three pharmacokinetic parameters: clearance (Cl), half-life (T1/2) and volume of distribution (Vd) are used to calculate the loading dose (Ld), maintenance dose (Md) and interval and duration of action of drugs. These are most useful when dealing with drugs with a narrow toxic/therapeutic ratio.

Most studies of drug kinetics refer only to profiles of plasma concentration following intravenous injection. Fortunately, most drug effects are closely related to plasma concentrations, the Ceff (see above). After single injection, plasma levels decline according to well-defined patterns. Abstract models representing one or more 'compartments' are frequently used to help predict duration effects.

Single compartment model

Here, the plot of plasma concentration (following single bolus i.v. injection) against time is a simple exponential curve which appears as a single straight line if the logarithm of concentration is taken. This indicates that the drug is distributed into a single compartment from which it is being metabolised by first order kinetics i.e. a constant fraction of drug present in plasma is eliminated per unit of time, the amount metabolised being proportional to the concentration of the drug in the compartment studied.

Two or more compartmental model

Most drugs do not behave as if they remain in a single compartment. After bolus i.v. injection into a 'central compartment' (e.g. plasma volume), they are rapidly re-distributed into one or more 'peripheral compartments', such as visceral organs with a large blood supply, muscle and fat. The concentration/time plot shows a rapid initial decline in concentration with a consequently short redistribution half life ( T1/2 alpha , measured in minutes). This is followed by a slower decline (or series of declines) in concentration due to the combined effects of elimination and continued re-distribution with a consequently long elimination half life (T1/2 beta ) measured in hours. Drugs falling within this pattern of disposition are sometimes described as 'short-acting' because of the rapid initial fall in plasma concentration during the distribution phase, e.g. thiopentone and fentanyl. They may, however, have 'cumulative' effects if administered repeatedly, due to a much slower elimination phase. Most drugs in current use show this pattern of distribution and elimination.

Volume of distribution

The volume of distribution (Vd) represents a theoretical 'volume' in which the drug distributes which would account for the concentration achieved in the plasma due to dilution after injecting a known dose of the drug. As many drugs bind to lipid and plasma proteins, this 'dilutional' volume calculated by dividing dose administered divided by concentration achieved (see below) may appear to greatly exceed total body water. Thus, Vd in a single compartment model can be easily calculated by dividing the dose injected (D) by the estimated initial plasma concentration (Ci, obtained by extrapolating the straight line in the concentration/time plot.

Vd is much more difficult to calculate in a two (or more) compartment model as the effects of continuing redistribution occur at the same time as metabolism and elimination. It is convenient to divide the Vd into two parts, the initial Vd (Vd init) and the steady state Vd (Vd SS) usually only obtainable by calculation after infusing the drug over a long period of time. In the case of opioids such as pethidine, Vd init may be only 50 litres whilst the Vd SS is over 200 litres.

Clearance (Cl)

If, as stated above the amount of drug metabolised is proportional to concentration, then

The constant is known as clearance and is the volume of plasma completely cleared of drug per unit of time and represents the best measure of the elimination capacity of the body. It is only a constant if first order kinetics apply to the metabolism of the drug because the body has excess capacity to metabolise the drug well beyond the clinical range. The units of clearance are volume per unit time and give no indication of the amount of drug metabolised. This will be dependent on concentration of the drug at the particular time. If Ceff is the desired concentration of the drug (in mg. per litre) required to produce an effect and Cl is the clearance (in litres per hour), then the amount metabolised in mg. per hour is equal to:

Equation 2 Cl * Ceff

Zero order kinetics is seen for a few drugs which follow a quite different pattern of elimination (e.g. phenytoin, ethanol and thiopentone in high doses). Above a certain plasma level, saturation of metabolic processes occurs. At or beyond this level, a constant amount of drug is eliminated per unit time regardless of plasma concentration.

Half life (T1/2)

This indicates the time it takes for the concentration (in the plasma) to fall by half. It does not necessarily mean that the amount in the body falls by half or that half is eliminated. As seen above, in the two compartment model the rapid initial fall in blood concentration means that the T1/2 alpha is only a few minutes, whilst the shallower curve of metabolism and elimination means that the T1/2 beta is many hours. Indeed, for many drugs, further half lives are calculated reflecting slow return of the drug from peripheral binding sites, often over many days. However, the plasma levels are usually so low that no clinical effect persists.

Half life is often utilised to predict 'wash-in' and 'wash-out' times, five half lives being taken to achieve wash-in or wash-out. This is particularly pertinent if the drug is commenced without a loading dose (see later). For a drug like digoxin with a T1/2 beta of 30 hours, it will take 150 hours to achieve steady state concentrations. Manipulation of the dose to achieve a desired effect must take into account this long wash-in time if inadvertent toxicity is to be avoided.

Trying to predict the duration of action of a drug by referring to it's T1/2 beta is fraught with difficulties. For a start, a drug may only achieve an effect at a specific concentration, thus even a slight fall will result in loss of effectiveness. On the other hand, receptor binding may be so avid that the half life is irrelevant to the duration of action (see later).

Calculation of loading dose (Ld)

In anaesthesia, it is often important to bring plasma concentration to therapeutically effective levels immediately. Rearranging Equation 1, a loading dose (Ld) is administered which is sufficient to fill the Vd to a desired drug level, normally the effective concentration, Ceff.

The level then falls, unless followed by the administration of the maintenance dose. In the case of an induction agent like thiopentone no further maintenance dose is given and anaesthesia continued with an inhalational agent. In the case of a more rapidly metabolised drug such as propofol, a maintenance regime can be used to continue anaesthesia entirely by the i.v. route.

As stated above, for more complex models it may be necessary to give a small Ld rapidly to take account of the Vd init and then a loading 'infusion' is given over an hour or so to take account of the Vd SS. If only the Vd SS was used to calculate the loading dose, toxic levels would occur initially.

Calculation of maintenance dose (Md)

Ceff is often achieved by means of continuous i.v. infusions (e.g. propofol and lignocaine), where the rate of drug administration (Md) equals the rate of elimination (see Clearance) :

Equation 4 i.e. Md = Ceff x Cl

If drug administration begins at maintenance dosage without a loading dose, it takes 5 half-lives to reach 96% of the steady-state plasma concentration. This usually prolongs the time taken for the drug to produce its optimal effect.

Example of calculation of loading and maintenance doses for lignocaine:

• Desirable effective concentration (Ceff) in plasma is 2 mg l-1 (1.5 - 6)
• Volume of distribution (Vd ) = 1100 ml kg-1 or 77 l (70 kg man)
• Clearance (Cl) = 9.2 ml min-1 kg-1 or 0.64 1 min-1 (70 kg man)

  Thus, from Equation 3:

Ld = Ceff x Vd

= 2 x 77

= 154 mg

From Equation 4

Md = Ceff x Cl

= 2 x 0.64

= 1.3 mg min-1

Thus, Vd determines the loading dose and Cl the maintenance dose. Cl is particularly important for drugs principally excreted via the kidney when administered to patients in renal failure. Loading dose will be the same or even higher than in patients with normal renal function, as Vd may be increased. Maintenance dose, however, will be markedly reduced (lower doses or longer dose intervals) in view of the much lower drug clearance. In congestive heart failure both Vd and Cl of lignocaine are reduced, necessitating lower loading and maintenance doses.

N.B. in clinical practise, dose regimes are normally calculated on a mg. per kg. basis (e.g. lignocaine 2 mg.kg-1 loading dose). The above calculations (equations 3 and 4) explain how these figures are derived.

The above calculations apply to drugs eliminated by first order processes, and the units used (e.g. mg, ml, minute) must be consistent between equations.

It is important to realise that there is a relationship between Vd, Cl and T 1/2 elim. i.e.

T 1/2 elim = Vd/Cl * 0.693

This relationship allows one to work out the third parameter when you know the other two. e.g. if you know the Vd and T 1/2 elim. you can calculate the clearance. It also shows that drugs with high volumes of distribution or low clearances (or a combination) tend to have long half lives (and vice versa).

The pharmacokinetic patterns resulting from metabolism and elimination, described above, result from chemical alteration (metabolism) and/or excretion of the drug.

This is carried out mainly in the liver. In general terms the liver converts active, lipid soluble drugs into inactive, water soluble drugs which are able to undergo renal or biliary excretion (lipid soluble drugs are filtered by the kidney but then immediately re-absorbed). Drugs already water soluble, such as digoxin and gentamicin, can be excreted directly. Obviously there are many exceptions to this general statement (see below).

Although a few drugs are metabolised by a single process, hepatic biotransformation is usually carried out in two stages:

• Phase I reactions (non-synthetic) are carried out by hepatic intracellular microsomal systems requiring NADPH, oxygen and cytochrome P-450. Examples are oxidation, reduction and hydrolysis. Drugs are usually prepared in this way for the second stage.
  
• Phase II reactions (synthetic) include conjugation with glucuronide, acetylation and methylation. These reactions render lipid soluble drugs water soluble, thus facilitating renal excretion.

Some of the above reactions, particularly microsomal oxidation and reduction, can be significantly affected by patient genetic makeup and enhanced by previous exposure to similar drugs. This latter is called enzyme induction. Classic examples are the chronic intake of phenobarbitone, rifampicin or alcohol which not only enhance their own metabolism but also metabolism of other drugs. Enzyme inhibition occurs with drugs such as metronidazole, cimetidine, erythromycin and sulphonamides.

Hepatic biotransformation is dependent upon both hepatic enzyme activity (HEA) and blood flow (HBF). For the relationship between clearance, hepatic blood flow and hepatic extraction ratio see the relevant section of the Powerpoint presentation.

• Low clearance drugs (e.g. aminophylline at 50 ml. min-1 in the 60 kg. patient) are more dependent on the level of HEA than HBF. Thus, substances which induce HEA (such as nicotine in this case) increase clearance whereas those drugs which inhibit HEA (such as erythromycin) decrease clearance and may lead to toxicity.
  
• High clearance drugs are those whose metabolism depends primarily on the level of HBF, such as opioids and beta blockers which have a clearance equivalent to HBF (1000 ml. min-1 in the 60 kg. patient). Clearance of these drugs are much more closely related to HBF than HEA, a fall in HBF reducing clearance. When given orally, these drugs undergo substantial 'first pass' metabolism, so blood levels are markedly increased if HBF is reduced. This, again, can cause toxicity.

Drugs like suxamethonium and mivacurium are metabolised by plasma enzymes (pseudocholinesterase which is produced in the liver). More rarely, enzyme systems in other organs, e.g. kidney or lung, are involved.

As stated above, there are many exceptions to the general rule that the liver converts active compounds into inactive ones:

• Active drugs are converted to compounds which are also active; e.g. diazepam is metabolised to desmethyldiazepam, which is equipotent to diazepam, but with a T1/2 b of 53 hours compared to 32 hours for diazepam. This is of importance in long-term administration.
• Inactive drugs are converted to active ones (e.g. triclofos into trichloroethanol).
• Active drugs are converted into toxic ones, e.g. pethidine to nor-pethidinic acid

Excretion of unaltered drug or its metabolites occurs mainly in the kidney, but also through the bile (e.g. vecuronium and buprenorphine) or through the lung (inhalational anaesthetics). If the plasma kinetics are highly dependent upon the renal excretion of non-metabolised drug (e.g. gentamicin), then the state of the renal function is important in determining duration of effect.

Pharmacodynamics is the study of mechanisms of drug action or 'what the drug does to the body'. Most drugs act upon receptors on the surface of cells in a highly specific fashion; very small modifications of the molecular structure of the drug may lead to a complete change of pharmacological activity. The study of this structure-activity relationship has led to the rational development of many new drugs in the past three decades.

A smaller proportion of drugs do not produce effects by interaction with specialised receptors. For example, the volatile anaesthetic agents are thought to act non-specifically upon the lipid component of cell membranes; magnesium sulphate exerts its purgative action by an osmotic effect within the gut; other drugs interfere with DNA, the genetic material.

These are situated on the surface of cells and mediate the action of natural chemical messengers upon intracellular mechanisms. An example is given in the section on anticholinergic drugs. It should be stated that the exact nature of transmitter-receptor interactions is not completely understood. The quantities of receptor molecules per gram of tissue is itself subject to complex regulation; e.g. thyroid hormone regulates the synthesis of beta receptors on heart cells.

The study of drug effects led to the discovery of different receptors naturally stimulated by the one transmitter; e.g. alpha and beta adrenergic receptors.

This may be classified as agonist, antagonist and partial agonist

• Agonists bind to receptors (affinity) and initiate activity (efficacy).
• Antagonists also bind to receptors (affinity) but have no activity (efficacy). They may displace agonists, e.g. naloxone has higher affinity for opioid receptors than morphine, displacing the latter at lower concentrations.
• Partial agonists often bind with greater affinity but have less activity (efficacy) than full agonists. The circumstances in which a partial agonist may act as an antagonist or an agonist depends on both the efficacy of the drug and the pre-existing state of receptor occupation by agonist.

For example, the high efficacy partial agonist opioid buprenorphine only acts as an antagonist if there is excessive agonist already occupying the receptor. Thus, in fentanyl overdose buprenorphine acts like naloxone and reverses respiratory depression. However, in the presence of small amounts of agonist it acts in an additive way, e.g. a patient in pain following an inadequate dose of pethidine gains additional pain relief if buprenorphine is added (or vice versa). In normal clinical use buprenorphine is such a high efficacy partial agonist that it is indistinguishable from other full agonist opioids such as morphine.

The low-efficacy partial agonist practolol almost always acts as an antagonist (i.e. a beta blocker). However, if there is zero underlying sympathetic activity (unlikely), the drug could act as an agonist (intrinsic sympathomimetic activity). Theoretically, there is a lessened risk of inducing profound bradycardia in a patient with very low underlying sympathetic activity (unlike propranolol which slows the heart rate further still).

The medium efficacy, orally active, partial agonist xamoterol lies in between these two extremes and is used in the treatment of mild to moderate cardiac failure. It can act as an agonist (i.e. like adrenaline) if the underlying level of sympathetic activity is low to moderate thus improving cardiac function. In severe heart failure the underlying sympathetic activity is high so it acts like a beta blocker. This leads to severe cardiac decompensation so the drug must be avoided in these patients.

To produce an effect, a drug must bind to a receptor, for that effect to disappear the drug must dissociate again. Thus:

In other words, as the drug concentration rises the receptor is occupied and an effect is produced, as it falls (due to distribution, metabolism and elimination) it becomes unoccupied and the effect diminishes. For the majority of drugs this is true, the continuing effect being determined by the drug level as predicted by pharmacokinetic data. However, a number of drugs do not act in this way and continue to occupy the receptor even though the level in the surrounding biophase is low. The ability to do this is determined by receptor 'avidity'. Buprenorphine, salmeterol (the long acting beta 2 agonist) and ondansetron are good examples of drugs with high receptor avidity. Although sharing almost identical pharmacokinetics with fentanyl, buprenorphine is only 50% dissociated from the opioid receptor at 1 hour versus 90% at 10 minutes with fentanyl. Obviously, buprenorphine is much longer acting. With salmeterol the longevity is probably due to exo-receptor binding of the lipophilic tail of the drug which allows it to repeatedly occupy the receptor.

N.B. the term 'receptor' may also appear in a physiological context referring to 'sensory receptors'; these are complex structures such as muscle spindles or retinal rods.

The activity and other properties of many drugs are related to chemical structure. New drugs are 'tailored' to fit therapeutic requirements. For example isoflurane and enflurane were designed to produce general anaesthesia based on the ether molecule, but with a lower solubility in blood, less irritant effects, and non-flammability.

Many definitions fall within this general heading, such as potency, efficacy, therapeutic index, and tolerance. Some of the useful concepts are better understood with the help of

The figure depicts the usual sigmoid shape of the log dose or log concentration versus effect relationship. Not all drugs have a graded response; some drugs produce an all-or-none type of response, such as suppression of dysrhythmias; similar plots may still be constructed, but in terms of frequency of occurrence of therapeutic effect in a population of individuals.

• Potency is related to the concentration or dosage which produces an effect (also 'affinity', in receptor theory).
• Efficacy is the maximal possible response which can be produced (also 'activity', in receptor theory).
• Tolerance is the decrease, over a period of time, in therapeutic effect produced by identical doses of the drug.
• Tachyphylaxis is the rapid development of tolerance.
• Therapeutic index or selectivity is the ratio between the dose producing undesired effects and the dose producing therapeutic effects. The median effective dose ED50 is that required to produce a certain intensity of effect in 50% of subjects. The median lethal dose LD50 is that which produces lethal effects in 50% of individuals (this evidently only applies to animal tests).

Adverse effects of drugs can be divided into type A and type B. Type A occur as much more pronounced effect of the drug at normal dosage, e.g. excessive bradycardia with a small dose of propranolol. Type B (bizarre ) reactions are not predictable and include immunological reactions such as penicillin allergy.

Plasma levels may now be measured for a large number of drugs. It is an expensive complement to drug therapy and should be used only if there are clear advantages. Blood samples must be taken at specified times after drug administration to be of any value. The indications are:

1. In some cases of treatment failure when there is a predictable relationship between plasma concentration and effects, e.g. anticonvulsants.
2. Use of drugs with a low therapeutic index or very serious side effects, e.g. lithium, gentamicin and theophylline.
3. In suicidal or accidental overdosage, e.g. paracetamol levels determine the management policy.

Caution about interaction of anaesthetic agents with other drugs is important. It may occur by competition for binding sites in plasma proteins, saturation of metabolic pathways in liver or plasma, inhibition of inactivating enzymes, or by physiological potentiation. Some of the best known examples are:

1. Ecothiopate eye drops used in glaucoma inhibit plasma cholinesterase prolonging the effect of suxamethonium.
2. Aminoglycoside antibiotics (e.g. gentamicin) have neuromuscular blocking properties of their own, potentiating the effect of pancuronium and other relaxants.
3. Monoamine oxidase inhibitors (e.g. iproniazid), by altering the metabolism of catecholamines and central analgesics, cause some patients to respond unpredictably to morphine or pethidine with hypotension, hypertension or coma. Response to inotropes and hypotensive agents is also unpredictable.

The physiology and pharmacology of the autonomic nervous system and circulatory control during anaesthesia.

The concept of the autonomic nervous system has resulted from anatomical and physiological studies carried out in the years since Langley first coined the term in 1898. The hypothalamus is the main site of integration of the autonomic nervous system, being a set of afferent and efferent nerve fibres and integrative neurones within the central nervous system. It represents the anatomical basis of regulatory reflexes which regulate visceral functions such as circulatory and temperature control, respiration, water balance, carbohydrate and fat metabolism. For this reason it has also been named the vegetative or involuntary nervous system. Studies in the past 20 years have shown that some regulatory functions can obey volitional control, suggesting cerebral cortical representation of some autonomic functions, e.g. blood pressure control.

The first demonstration of the chemical nature of neurotransmission was made by Otto Loewi (1921) whilst studying the vagal innervation of the heart. The concept of a drug receptor has also emerged from experimental work in autonomic structures, on finding that different drugs may mimic or block different effects of the same natural neurotransmitter at different end organs. Traditionally, the autonomic nervous system has been divided into two anatomically and functionally distinct efferent divisions: the sympathetic and the parasympathetic.

Originating mainly in the hypothalamus, it emerges from cells in the intermediolateral columns of the thoracic and lumbar segments of the spinal cord (T1-L2) as the myelinated, white rami communicantes ('B fibres') which hitch a ride in the anterior nerve roots. They supply the pre-ganglionic innervation of the para-vertebral chain of sympathetic ganglia and abdominal plexuses. Synaptic transmission in the ganglia is mainly cholinergic with both nicotinic and muscarinic as well as dopaminergic receptors. The unmyelinated, grey rami communicantes (C fibres) leave the ganglia as post-ganglionic neurones to rejoin peripheral nerves and supply the end organs with noradrenergic nerve endings.

The pre-ganglionic neurones receive excitatory and inhibitory synaptic inputs from the spinal cord, brain stem and other higher centres. There is a constant background discharge in these neurones which is referred to as underlying 'sympathetic tone'. This is responsible for maintenance of tone in the smooth muscle of both capacitance and resistance blood vessels, thereby maintaining blood pressure.

Since the B fibres may synapse immediately in the sympathetic chain or go up or down the chain to a higher or lower level (including the adrenals, themselves modified sympathetic ganglia), the sympathetic system tends to respond to stimuli as an integral system. All nerve endings discharge noradrenaline, with two known exceptions: sweat glands and blood vessels in voluntary muscle, which are cholinergic.

This has two main outflows from the central nervous system, one arising with cranial nerves and the other with sacral roots. Pre-ganglionic neurones are much longer than their equivalents in the sympathetic system but are also cholinergic. The ganglia are situated very close to or within the effector organs (e.g. Auerbach plexuses in the intestine). The post-ganglionic neurones are very short and release acetylcholine. The parasympathetic system tends to respond to stimuli in a more localised fashion. There is no evidence of parasympathetic tone regulating blood pressure.

Most viscera are innervated by both sympathetic and parasympathetic systems; the effects of stimulation are usually antagonistic (e.g. sympathetic accelerates the heart, parasympathetic slows it, other examples being the pupil and bronchial smooth muscle). A few effectors receive innervation only from the sympathetic system, such as sweat glands and 'resistance' blood vessels in the skin and gut.

The description just given is an oversimplification of a very complex system both anatomically and functionally. Ganglia for example, far from being simple relay stations, are neuronal networks with more than one neurotransmitter involved. In addition to the neural pathways and neurotransmitters described above, there are other fibres innervating the same end organs which release other chemicals; some have a well-established role, but many of the newly described substances do not exhibit uniform effects in different species, and the relevance to human physiology is hypothetical. Examples are bradykinin, 5-hydroxytryptamine, enkephalin, substance P and vasoactive intestinal peptide (VIP).

The natural transmitters adrenaline, dopamine and noradrenaline are often used therapeutically. A large number of synthetic drugs have been studied which mimic or block the effects of the natural transmitters at particular locations, depending upon the type of receptor present. Accordingly, they are designated agonists, partial agonists or competitive antagonists (see above).

The neurotransmitters described in the autonomic nervous system are also present in the central nervous system. Consequently, drugs prescribed for peripheral effects should always be considered to have central effects as well, their state of ionisation in plasma dictating whether they cross the blood-brain barrier.

There are three types of peripheral cholinergic receptors, and there may be more within the central nervous system:

• Nicotinic neuromuscular cholinergic receptors occur at the neuromuscular synapse of voluntary muscle. It is named 'nicotinic' after nicotine was found to be the first drug having agonist effects. Stimulation of this receptor produces depolarisation and contraction of voluntary muscle.
• Nicotinic ganglionic cholinergic receptors are also present at the synapses of both sympathetic and parasympathetic ganglia, and are also excitatory. The specific competitive antagonist drugs for this receptor are called ganglion-blockers and were the first drugs used to lower blood pressure in the early 1950's. Their effect is to reduce traffic of sympathetic impulses which maintain tone in vascular smooth muscle. This class of drug is not used in ambulatory patients because of postural hypotension, but pentolinium is still used during anaesthesia to lower blood pressure. The neuromuscular blocker tubocurarine has marked ganglion-blocking properties. Vecuronium and atracurium are much more specific for neuromuscular nicotinic receptors.
• Muscarinic receptors occur at parasympathetic nerve endings; muscarine mimics the action of acetylcholine at this receptor (see below).

Types and structure

Three anticholinergic drugs are in common usage. Atropine is a racemic mixture of D and L hyoscamine of which only the l form is active. Scopolamine is L-hyoscine and differs from atropine by the addition of an oxygen bridge (D-hyoscamine is much less active). Glycopyrrolate is a synthetic anticholinergic and possesses a quaternary ammonium structure which means that it is a water soluble, ionised drug at body pH, unlike atropine and scopolamine which are non-ionised and lipid soluble. Glycopyrrolate thus has limited membrane (e.g. gut, placenta and brain) penetrating potential. In comparison with atropine and scopolamine it is poorly absorbed orally, does not cause fetal tachycardia and does not interfere with central nervous system (CNS) cholinergic function.

Mechanism of action

These drugs are competitive antagonists (almost exclusively) at the muscarinic receptor with little effect at nicotinic sites. Their broad range of actions throughout the CNS and autonomic nervous systems are mediated via 3 stages:

1. receptor binding
2. activation of guanidine nucleotides (G proteins).
3. transduction via G proteins to produce intracellular effects on adenylyl cyclase, phospholipase C and ionic flux of calcium and potassium across cell membranes.

There are 3 well defined subtypes of muscarinic cholinergic receptors in humans, denoted M1 to M3, each encoded by a unique gene and differing in their amino acid sequence. M1 receptors are mainly found in the CNS, M2 in the cardiovascular system and M3 in secretory glands (e.g. salivary). All three drugs are non-specific antagonists at these receptors, but atropine has a two fold preference for M1 (CNS) receptors. M3 receptors (glandular secretion, blockade producing a reduction in salivation) are much more sensitive than M2 receptors (cardiac, blockade causing an increase in heart rate). Thus, for all three drugs, a greater dose is needed to prevent bradycardia when compared to reducing salivation.

Effects

In clinical practise, a 'normal' parenteral dose of scopolamine (0.4 mg. per 70kg) does not cause tachycardia whereas an equivalent dose of atropine (0.6 mg) and glycopyrrolate (0.4 mg.) does. Effects on reduction of secretions and intestinal motility are similar. Other major effects are summarised in table 2.1

Pharmacokinetics

Both atropine and scopolamine are well absorbed by the oral route whereas glycopyrrolate is not. Oral atropine has adequate M2 effects, thus preventing bradycardia, whereas oral glycopyrrolate is only effective in drying secretions (M3). Scopolamine is rarely used orally in anaesthesia but is now available in a transdermal preparation which may prove useful for prevention of motion exacerbated vomiting in ambulant patients following day surgery. All three drugs are well absorbed by the i.m. route. After i.v. administration, the onset of atropine is much faster with a shorter duration of action when compared with glycopyrrolate. This is due to atropine's much higher Vd, presumably as a result of increased lipid solubility.

Uses

Atropine is used as a premedicant in those cases where excess salivation and bronchial secretion is problematical, e.g. babies and patients undergoing upper GIT or respiratory endoscopy. During anaesthesia it is used to reverse or prevent bradycardia, particularly associated with anticholinesterases such as neostigmine which are used for reversal of residual neuromuscular blockade (q.v.)

Scopolamine is still used for premedication in combination with papaveretum where it causes sedation, anti-emesis and reduction in salivation.

Glycopyrrolate is used exclusively by the parenteral route, due to it's poor absorption from the GIT. Uses are similar to atropine but the lack of CNS side effects and better matched onset and offset time with neostigmine makes it the agent of choice in reversal of neuromuscular blockade, especially at extremes of age. It is also preferred in obstetrics for its lack of propensity to cause fetal tachycardia.

There are two main types of adrenergic receptors, alpha and beta, which may co-exist in the same effector organ. Subtypes of these receptors, alpha-1 and 2 and beta-1 and 2, have been identified with the advent of specific blockers and antagonists. They are distributed throughout the body, including the central nervous system, and both types are activated by noradrenaline and adrenaline. Activation of alpha and beta receptors often has opposite physiological effects upon smooth muscle, such as in bronchi and in arterioles, a poorly understood phenomenon. Possibly regulation of synthesis of either type of receptor determines the relative effect of natural transmitters.

Classic examples of location of beta receptors are the heart (excitatory), uterus (inhibitory), bronchi (dilatation), platelets (aggregation), and brown adipose tissue (heat production). Examples of alpha receptor location are the eye (mydriasis), skin and visceral arteries (constriction), veins (constriction), and bladder sphincter (contraction). Alpha receptors have been identified in many structures without a complete understanding of their physiological role; such as the spinal cord, associated with pain pathways, in the brain stem associated with the control of blood pressure and temperature, and in human platelets.

Understanding of intracellular mechanisms mediating the step between activation of receptors and cellular response is still poor. Cyclic AMP is one intracellular mediator of excitatory phenomena triggered by beta agonists.

• Alpha blockers are no longer used as hypotensive agents in ambulatory patients because of postural effects and reflex tachycardia; phentolamine is used in cardiac surgery to lower blood pressure by a reduction in systemic vascular resistance (SVR).
• Alpha-1 agonists such as phenylephrine are used to increase SVR and BP during cardio-pulmonary bypass.
• Alpha-2 agonists such as clonidine, pre-synaptically inhibit the release of noradrenaline at adrenergic neurone endings. Interestingly, these agents have been found to produce anti-nociception when applied spinally and to augment the effect of general anaesthetics. A more specific alpha-2 agonist, dexmetotomidine, reduces anaesthetic requirements by up to 90%.
• Beta blocking drugs have been extensively studied in recent years because of their application to long-term management of high blood pressure, dysrhythmias and anginal pain. The specificity of drugs for beta-1 and beta-2 receptors is far less marked than that seen between alpha and beta receptors. Asthmatic patients should never be given beta blockers; even drugs claimed as beta-1 specific may precipitate fatal bronchospasm.
• Beta agonists are widely used as inotropic agents (e.g. isoprenaline), as bronchodilators in asthma (salbutamol) and to arrest labour in obstetrics (ritodrine). Dobutamine is a synthetic beta agonist used to increase cardiac output in intensive care patients. Salbutamol, claimed to be beta-2 specific, may cause severe tachycardia due to the presence of beta-2 receptors in the atria. There is evidence that beta receptors are also present in the central nervous system, such as medullary neurones regulating respiration and sympathetic tone.
• Alpha and beta effects: some synthetic agonist drugs have both alpha and beta effects like the natural transmitters. They may act indirectly by promoting the release of noradrenaline (e.g. ephedrine) or by direct action on the receptor (e.g. metaraminol). Labetalol has both alpha 1, beta 1 and 2 blocking activity.

Dopamine is the immediate precursor of noradrenaline, the step being catalysed by dopamine-beta-hydroxylase (present within the microvesicles of noradrenergic nerve terminals). Groups of central neurones lacking the enzyme secrete dopamine, such as those found in pathways controlling voluntary movement (depletion causes Parkinson's disease), and vomiting reflexes. Peripherally, dopamine receptors have been identified in the kidney (vasodilatation) and in the carotid body chemoreceptor (inhibition).

Dopamine infused at rates of 1 - 3 ug kg-1 min-1 ('renal dose') specifically increases renal blood flow; at higher doses it also has beta, and finally alpha effects. Examples of synthetic dopamine agonists are apomorphine (used rarely to induce vomiting) and bromocriptine used in Parkinson's disease. Dopamine blockers are used in psychiatric disorders (haloperidol), in neuroleptic anaesthesia (droperidol), and as antiemetics (droperidol, prochlorperazine and metoclopramide).

• DA 1 receptors mediate relaxation of renal blood vessels and gut smooth muscle
• DA 2 receptors mediate presynaptic inhibition of dopamine and noradrenaline

Other systems

There are other transmitter systems in the autonomic system, as stated earlier. A number of drugs related to these systems are finding a place in clinical practice, such as ketanserin (5-HT blocker), in the management of certain types of chronic limb pain and ondansetron and granisetron (5- HT 3 blockers) in emetic chemotherapeutic regimes and post-operative nausea and vomiting.

Many other drugs act upon autonomic functions without direct receptor interaction. They are mostly hypotensive agents:

• The action may be predominantly central, such as reserpine, which drops blood pressure by reducing sympathetic 'tone' centrally.
• The action is at the presynaptic end of adrenergic terminals, by interfering with catecholamine synthesis (a methyl-dopa), storage (reserpine), release (guanethidine) or re-uptake (desipramine)
• The tricyclic antidepressants on the other hand block re-uptake of noradrenaline at the adrenergic neurone ending, thus enhancing it's effect.
• The enzymatic breakdown of neurotransmitter may be inhibited, enhancing the local concentration of transmitter. Monoamine oxidase inhibitors slow down the breakdown of catecholamines, both peripherally and centrally; the latter effect may be beneficial in depression. Neostigmine acts by a similar mechanism on cholinergic terminals by inhibiting cholinesterase; it does not cross the blood-brain barrier. Another cholinesterase inhibitor, physostigmine, crosses easily into the brain; it is thus able to reverse central depressant effects of overdosage with atropine, hyoscine or tricyclic antidepressants.
• Finally, there is a group of drugs which act directly upon the smooth muscle cell. Examples are sodium nitroprusside and nitro-glycerine whose hypotensive effects are initially mediated by relaxation of venous smooth muscle and at higher dosage by arteriolar dilatation as well. Hydrallazine also acts directly upon smooth muscle, with preference for the arteriolar side. Papaverine and theophylline are other examples of smooth muscle relaxants. The latter acts upon intracellular cyclic-AMP mediation of beta effects.

Having discussed the physiology and pharmacology of the autonomic nervous system it is appropriate to look at how these systems work in collaboration with peripheral control during anaesthesia and surgery. It is obvious that safely conducted anaesthesia requires:

• Sound knowledge of circulatory physiology and pharmacology
• Maintenance of adequate cardiac output and perfusion pressure
• Provision of normoxia and adequate oxygen delivery, normocapnea and adequate analgesia
• Minimisation of heat loss by maintenance of a warm operating environment (22-23 deg.C), use of warming blankets, warmed fluids and humidification of the inspired gas.

Theoretical aspects of Circulatory Control

Pressure, flow and resistance

The Poiseuille-Hagen equation describes the factors which determine flow of a homogenous Newtonian fluid through a non-distensible cylindrical tube:

Blood flow = Perfusion pressure * radius^4 * pi / length * viscosity * 8

-the resistive component above is determined by the radius of the blood vessel and viscosity of the blood. Radius (being to the power of 4) therefore has a dramatic effect on resistance. Blood is a non-Newtonian fluid due to the presence of suspended cells (red blood cells, leucocytes and platelets). This results in non-linearity of viscosity (the Fahreus-Lindqvist effect) whereby it is effectively reduced as the radius of the vessel diminishes. This tends to counteract the effects of increased resistance as the vessel narrows.

Thus, in simpler terms:

Blood flow to a tissue is proportional to pressure difference divided by resistance,

(cf Ohm's law of electrical resistance where V=IR or I = V/R)

Cardiac output (Q), blood pressure (BP) and systemic vascular resistance (SVR): principles of peripheral circulatory control

Cardiac output is the product of stroke volume and heart rate. Stroke volume is determined by preload (mainly resulting from venous return), inherent contractile state (determined by sympathetic stimulation and endogenous (or exogenous) catecholamines and afterload (resistance of the circulation to ejection of blood). Again, rearranging the equation above, BP (V) is the product of Q (I) and SVR (R). Frequent measurement of BP is fundamental to good anaesthetic practise. However, It is clear from the above that a fall in blood pressure may be due to either a fall in Q, a fall in SVR or a combination of the two.

During surgery and anaesthesia, a fall in Q results from:

• decreased preload and thus stroke volume resulting from hypovolaemia (absolute, due to blood loss or relative or from vasodilation due to anaesthetic agents).
• decreased contractility from the depressant effects of anaesthetic drugs or inadvertent hypoxaemia
• increased afterload due to peripheral vasoconstriction, often occurring in association with relative hypovolaemia and increased sympathetic tone

Since maintenance of Q (and oxygen delivery) is of such crucial importance, it is surprising that methods to measure Q directly are not routinely used during anaesthesia. Newer techniques of non-invasive measurement will assume increasing relevance in the future, including oesophageal and transtracheal Doppler ultrasound and thoracic bioimpedance techniques. These methods have a good correlation with more invasive (and inappropriate) techniques such as thermal dilution. BP must be maintained by volume replacement to restore Q.

A fall in SVR usually results from the vasodilator effects of the agent (see below) either directly (as with isoflurane and propofol) or indirectly due to histamine release (such as d-tubocurarine).

The reflex response to hypovolaemia and a falling venous pressure is mediated by the 'low pressure' cardiopulmonary baroreceptors situated in the vicinity of the superior vena cava and right atrial junction. This results in an increase in sympathetic tone and peripheral vascular resistance with arteriolar and venular constriction (Q is actually less than normal). The main site of increase in SVR is the arterioles and to some extent the small arteries and the precapillary sphincters. Although this may not be marked in some organ beds (eg. cerebral), it is in skin and muscle.

The 'high pressure' baroreceptor reflexes are activated by sensors located in the arch of the aorta and respond to a falling arterial pressure (from any cause) by increased sympathetic activity. During anaesthesia, this reflex increase in SVR (e.g. to hypovolaemia) may be effectively blocked by inhalational agents (halothane > enflurane> isoflurane) and i.v. induction agents (propofol> thiopentone> etomidate) all of which attenuate these central neural reflexes that stabilise BP. In addition, direct vasoconstricting effects on the peripheral circulation (e.g. to circulating catecholamines) are depressed. Nitrous oxide, benzodiazepines, ketamine, pancuronium and opioids such as fentanyl (not morphine and pethidine) have much less effect.

The elderly and volume depleted patients are much more sensitive to the haemodynamic depressant effects of anaesthesia and often respond with a precipitate fall in BP. It is thus crucial to adequately fluid resuscitate hypovolaemic patients prior to anaesthesia, avoiding where possible agents which are particularly vasodepressive. Conversely, high doses are needed to produce profound hypotension during anaesthesia in the young, normovolaemic patient who is surgically stressed.

Longer term effects on BP and SVR are humorally mediated through release of agents with marked vasoactive properties such as adrenaline, noradrenaline, vasopressin, prostaglandins, kinins and angiotensin II. In general, the larger arterioles (first and second order) are under predominantly neural influence whilst the smaller arterioles (third and fourth order) respond mainly to humoral (and direct vasodilator) influences.

Regulation of blood flow through capillary beds is mainly concerned with the supply of adequate oxygen and nutrients to cells requiring them. At any one time, only small sections of the capillary bed are perfused. As metabolism continues to proceed, hypoxia and hypercarbia supervene, together with the production of metabolic products such as adenine, hydrogen ions, potassium and lactic acid. These substances tend to cause relaxation of the precapillary sphincters to allow blood flow to be re-established. Thus, given an adequate supply of blood to the tissue bed (dependent also on remote factors) it can then be appropriately distributed by local factors (see also section on endothelium).

The role of the intact endothelium in the production of vasodilator substances such as the endothelial derived relaxing factor (EDRF, now known to be nitric oxide) and prostacyclin (PGi2), is substantial. Nitric oxide (NO) is formed from l-arginine under the influence of the enzyme NO synthase. NO is a low molecular weight volatile free radical (unpaired reactive electron in the outer electron shell which can thus combine covalently with other molecules). It combines with and activates guanylate cyclase in vascular smooth muscle cells: causing vasodilation, inhibition of platelet aggregation and adherence of neutrophils. NO synthase exists in two main forms. In basal conditions it is present as a calcium and calmodulin dependent enzyme which can be activated by vasodilators such as acetylcholine, adenosine, histamine and bradykinin. This form is controlled by cell-surface receptors. A calcium independent form is produced by sepsis and cytokines, which once 'turned on' is no longer under control of cell surface receptors. Thus, excess NO synthase may cause the hypotension and myocardial depression of septic shock.

NO and PGi2 are produced in response to potentially vasoconstricting factors, such as thromboxane (TxA2) and adenosine diphosphate (ADP) produced from platelets following vascular damage and endothelial injury. These vasodilatory factors, together with fibrinolytic agents such as tissue plasminogen activator (TPA) limit stasis and thrombosis following endothelial damage. In their absence, coagulation and thrombosis occur, the endothelium and distal tissue become hypoxic and events are set in motion which can lead to irreversible cellular damage.

To complicate the issue still further intact endothelium also produces vasoconstrictor substances such as endothelium-derived contracting factors (EDCF) and endothelin. The latter is a 21 amino acid peptide which is has greater vasoconstricting activity than any other hormone known. It is also responsible for the metabolism of renin and angiotensin 1 to produce angiotensin II, another potent vasoconstrictor.

Additional factors of relevance during anaesthesia include the effect on the endothelium of smoking, ageing, arteriosclerosis and endothelial changes due to hypertension and hyperlipidaemia. It would appear that in these circumstances the endothelium is much more sensitive to contracting substances as well as being less able to prevent the formation of thrombus due to tissue damage.

Conclusion

It is clear that circulatory control during anaesthesia, especially if compounded by the effects of surgery, hypovolaemia and sepsis, is extremely complex. The major points of note involve the provision of adequate anaesthesia and analgesia without compromising cardiac output and oxygen delivery, together with maintenance of normovolaemia.