TRAINEES FORUM: PREPARING FOR THE PRIMARY FRCA Editor Dr J A Langton, Derriford Hosptial, Plymouth

Total Intravenous Anaesthesia (TIVA) II:

Pharmacodynamics of intravenous anaesthetics
Dr K J Anderson and Professor G N C Kenny, Glasgow Royal Infirmary

harmacokinetics describes the distribution and elimination of

P drugs by the body, in particular the relationships between drug
concentration and time. The relationship between concentration
actual brain concentration of drug, the closest we have in animal
models, is sampling from the cerebral venous system. Even if we
could measure direct brain concentration, to relate concentration to
and response is known as pharmacodynamics, and describes what effect properly, it would be necessary to know the exact regional
the drug does to the body. For intravenous anaesthetics the plasma concentrations or even receptor concentrations where the drug
is not the site of drug effect. exerts its effect.
Therefore, we must rely on the clinical effects produced by the
The relationship between plasma drug, to describe the relationship between the blood concentration
concentration and effect
and brain concentration or effect. The commonly used end points
It is apparent during induction and emergence from anaesthesia, of clinical effect are loss of verbal contact or lack of response to sur-
that the plasma concentration (Cp) does not always correlate, with gical or noxious stimulus. Assessment of this is usually quantified
the clinical effects of anaesthetic drugs (Figure 1). by interval scales such as the Objective Observer’s Assessment of
For instance, during induction (at point A) the effect site con- Sedation Scale. Other more convenient measures of anaesthetic, or
centration (Ce) lags behind the plasma concentration, hence the more specifically hypnotic effect have been used, predominantly
plasma concentration is higher than the effect site concentration. measures of cortical activity. The most commonly used of these are
Again during emergence (at point B) the effect site concentration processed EEG signals such as Bispectral Index (BIS) and 95%
lags behind the reduction in plasma concentration, however at this Spectral Edge Frequency (SEF95), or by Auditory Evoked Potentials
point the effect site concentration is higher than the plasma con- (AEP). These monitors of ‘anaesthetic depth’ are useful surrogate
centration. Since time is required for the blood and brain concen- markers of the effect site concentration. They make it easier to
trations to equilibrate it has been suggested that when titrating the relate plasma concentration to effect because they are objective
drug to effect, it would make more sense to consider the effect site rather than subjective, and also they are measured on a continuous
concentration which is related to that clinical effect. The mathe- scale.
matics of the equilibration of anaesthetic drug plasma concentra-
tion and effect site concentration are described by the pharmacoki- Effect site model
netic rate constant ke0. If the clinical effects of the drug are measured with respect to the
plasma concentration, then the apparent rate of drug flow into
Effect Site
and out of the biophase can be mathematically modelled. Thus,
The principal effects of i.v. agents in which anaesthetists are inter- by the addition of another (effect site) compartment to the three
ested are the sedative and hypnotic effects. Thus the site at which compartment pharmacodynamic model (Figure 2), the expected
the drug exerts these effects (termed the biophase or effect site) is time course of the clinical effect for a plasma concentration can be
the brain. Unfortunately it is impossible to measure the calculated.
Figure 1 The relationship of effect site to plasma propofol concentra- Figure 2 The three compartment pharmacokinetic model with added
tion during induction and emergence with TCI effect site compartment (Ce)

Bulletin 17 The Royal College of Anaesthetists January 2003 831

TRAINEES FORUM: PREPARING FOR THE PRIMARY FRCA Total intravenous anaesthesia II Anderson KJ, Kenny GNC

Movement of drug to and from the central compartment (plasma) thiopentone and two minutes 30 seconds for propofol respective-
is considered a first-order process, and the effect compartment is ly), the concentration gradient for equilibration of plasma
defined as being negligibly small in comparison to compartments thiopentone with the effect site is continually reducing faster.
C1–C3. Therefore, the addition of this compartment has no effect Hence if we plot the plasma concentration and the effect site con-
on the concentration of the other compartments. For this reason centration of the bolus doses, then the peak effects are at the point
some people consider movement from the effect site compart- when the falling plasma concentration crosses the rising effect site
ment to be drug eliminated from the model. Accordingly, the rate concentration (Figure 4).
constants for delivery to and disposal from the effect site are called
Figure 3 The effect that t 1/2ke0 has on time to equilibrate the effect
k1e and ke0 respectively (kχ0 is the time constant for elimination of
site with a constant plasma concentration of thiopentone and propofol
drug from the model from any compartment named χ). respectively
Furthermore, since the effect compartment is negligibly small, k1e
is so much smaller than ke0 as to be disregarded. Therefore, ke0
characterises the equilibration of plasma concentration with drug
effect. This allows calculation of the effect site concentration.
Numerous studies have calculated ke0 for various anaes-
thetic drugs, the value seems to depend on the patient popu-
lation and the surrogate marker of effect used, e.g. BIS or AEP.

ke0: Describing drug delivery to the effect site

By definition, ke0 is the pharmacokinetic rate constant (for the
Michaelis Menton equation), which describes the rate of equilibra-
tion between the plasma concentration and effect site. Perhaps a
conceptually easier term is the t 1/2ke0. That is the time it takes for
half of the equilibration to take place between the plasma concen-
tration (Cp) and the effect site concentration (Ce). Importantly,
the concept has been incorporated into the Diprifusor package,
which now can calculate and display effect site concentration also.
This helps illustrate to practicing anaesthetists the difficulty in
relating calculated Cp, or indeed measured Cp to effect. It can be
seen from the relationship of Cp to Ce that the same calculated
plasma concentration could be associated with greatly different
effect site concentrations. It becomes obvious that the clinical
effects depend on how long the blood and effector site have been
allowed to equilibrate.
The value of this new knowledge can be illustrated if we com-
Figure 4 The relationship of plasma concentration and effect site con-
pare drug disposition and effect following induction of anaesthesia
centration following a 2mg·kg-1 bolus of propofol and a 5mg·kg-1 bolus of
with 2mg·kg-1 of propofol with 5mg·kg-1 of thiopentone. Simply by
thiopentone
observation, we know that the onset of anaesthesia is slower with
propofol than with thiopentone. Computer pharmacokinetic simu-
lation confirms this, predicting the peak effect site concentration of
the thiopentone bolus at 1.4 minutes and the propofol bolus at four
minutes. Firstly, if we consider the effect of the t 1/2ke0, we can see
that for a constant plasma concentration of thiopentone that the
effect site will equilibrate more quickly than propofol (Figure 3).
The t 1/2ke0 of propofol is slower at 2.6 minutes compared to 1.2 min-
utes for thiopentone.
However, in real life, drug redistribution and elimination is
also taking place concurrently as the plasma effect sites equilibrate.
Since thiopentone is initially redistributed more quickly than
propofol (t1/2 immediately bolus dose is one minute ten seconds for

832 Bulletin 17 The Royal College of Anaesthetists January 2003

TRAINEES FORUM: PREPARING FOR THE PRIMARY FRCA Total intravenous anaesthesia II Anderson KJ, Kenny GNC

Table 1 Relationship between Target and Effect site CP50 and CP95 of
propofol for clinical end points. This is without any other sedative or
opioid medication. (From Irwin MG et al. Anaesthesia
2002;57:242–248).

CP50 (µg.ml-1) CP95 (µg.ml-1)

Clinical endpoint
Effect Target Effect Target

Loss of Verbal Contact 2.7 3.9 3.8 5.4

Loss of purposeful
response to surgical 4.5 5.6 6.4 7.4
stimulus

Figure 5 The dose response curve for probability of anaesthesia (lack

This is because drug will always move down its concentration gradi- of response to noxious stimulus) versus calculated effect site and plas-
ent, that is from plasma to brain when the plasma concentration is ma concentrations of propofol. (Adapted from data by Irwin MG et al.
greater than the brain concentration. When plasma concentration Anaesthesia 2002;57:242–248)
falls below brain concentration, drug movement reverses.

Population pharmacodynamics
MAC and Cp50
Most anaesthetists are familiar with the term MAC (minimal alveo-
lar concentration of an inhalational anaesthetic which prevents gross
purposeful motor response in 50% of unpremedicated patients to
skin incision). A similar concept has been developed for intravenous
agents, and is referred to as the effective concentration 50 (EC50 or
Cp50). This is the plasma concentration (Cp) that prevents gross pur-
poseful motor response in 50% of unpremedicated patients to skin
incision. It is a preferred term to effective dose 50 (ED50), which
relates to the dose of drug that prevents gross movement in 50% of
the population to skin incision. In a population, for the same dose Similarly increasing the dose of midazolam given prior to induction
of drug, there will be variation in the blood concentration achieved of anaesthesia with propofol has been shown to reduce the ED50, and
owing to pharmacokinetic variation. This introduces a further level increase the proportion of patients successfully anaesthetised with a
of uncertainty not present when relating plasma concentration to fixed calculated blood concentration (Figure 7).
effect. Both Cp50 and MAC are useful for comparing the potencies of Interestingly for opioids it is becoming clear that there is a ceiling
anaesthetic agents within their respective class, but by definition only to their effect on Cp50 and Cp95, and no matter how high the opioid
provide adequate anaesthesia for 50% of the population. A more concentration, some hypnotic is necessary to prevent movement to sur-
useful concept clinically is Cp95, which is the effective concentration gical incision (Figure 6). Therefore, it can be surmised that opioids
for 95% of the population. It is likely that expressing Cp50 and Cp95 may reduce the magnitude of the surgical stimulus but do not provide
in terms of calculated effect site concentration (Ce50 and Ce95) will hypnosis. It should be noted that although propofol concentration can
give even more clinically useful information (Table 1). From this be reduced in the presence of remifentanil, if the plasma concentration
data, dose-response curves for propofol plasma and effect site con- of remifentanil is greater than around 1.5 ng.ml-1, then the patient is
centration can be constructed (Figure 5). likely to become apnoeic and require artificial ventilation (Figure 6).
The interaction between opioids and propofol for total intravenous
Pharmacodynamic interactions
anaesthesia has been studied extensively. Computer pharmacokinetic
Similar to the effect on MAC of volatile anaesthetics, concomitant simulations have been performed to calculate the optimum combina-
administration of sedative premedication, benzodiazepines, opioids, tion that provides adequate anaesthesia with minimal time to return of
nitrous oxide or indeed α2 adrenoceptor agonists such as clonidine all consciousness when the infusions are ended. The optimum combina-
reduce the Cp50 for intravenous agents. This effect exhibits a dose tion changes depending on how long the infusions have been running
response, for example, by increasing the concentration of opioid the (the context sensitive half-time) and the opioids used.
MAC and Cp50 are reduced further (Figure 6).

Bulletin 17 The Royal College of Anaesthetists January 2003 833

TRAINEES FORUM: PREPARING FOR THE PRIMARY FRCA Total intravenous anaesthesia II Anderson KJ, Kenny GNC

Figure 6 The plasma remifentanil versus plasma propofol concentra- Systemic disease
tion associated with no movement to surgical stimulus in 95% of In contrast to the well characterised effects of systemic disease,
patients (Cp95). Superimposed is the level of remifentanil at which
especially renal and hepatic disease, on the pharmacokinetics of
patients may be expected to breath spontaneously (from Mertens MJ.
intravenous hypnotics and opioids, there has been surprisingly
PhD Thesis. ISBN 9090156410. Chapter6)
very little work performed to examine their relative potency in
patients with systemic disease. It has often been assumed that
patients with significant disease would require less anaesthetic.
This variation could be as a result of increased central nervous sys-
tem sensitivity to the drug (pharmacodynamic), or an increased
free fraction of drug secondary to reduced plasma protein binding
(subtle pharmacokinetic changes). An increased clinical effect for
a defined blood concentration has not been shown in disease
states. In fact, patients with hepatic cirrhosis have been shown to
waken at the same plasma concentration of propofol as healthy
controls. Furthermore surprisingly little difference in plasma pro-
tein binding of propofol has been found in patients with hepatic
and renal disease.
Figure 7 Clinical interaction of midazolam and propofol. The propor-
Titrating anaesthesia to stimulus
tion of patients losing verbal contact with a fixed target controlled infu-
sion of propofol at 3 mg·ml-1 after varying doses of midazolam (Tzabar Despite the fact that variation in Cp50 for intravenous anaesthetics
Anaesthesia 1996;51:536–538) has not clearly been shown to change with disease there is significant
patient variability to the pharmacodynamic effects of intravenous
anaesthetics in health, and due to aging. Obviously, the likelihood
of response will depend on the depth of anaesthesia, effect site con-
centration and the degree of surgical stimulation. So in practice the
depth of anaesthesia will need to be titrated to the surgical stimula-
tion. To complicate matters further the requirements differ with the
effect considered. For example the opioid concentration necessary
for intubation is higher than for incision. It becomes obvious that,
just as for inhalational anaesthesia, the anaesthetic depth should be
individualised by titrating the different components of TIVA to the
individual patient and the surgical stimulus at that moment. A
sound knowledge of the pharmacokinetic and pharmacodynamic
principles of intravenous anaesthetic agents will make appropriate
Pharmacodynamic variability drug selection and administration possible.
Age
Increasing age has been shown to reduce the Ce50 for propofol, Further reading
showing increased sensitivity of the elderly to the effects of propo- Anaesthesia 1998;53 (supplement 1):1–86.
fol. Importantly they exhibit greater haemodynamic effects also.
Padfield NL. Total Intravenous Anaesthesia. Butterworth-
Interestingly the ke0, hence plasma effect site equilibration has been
Heinmann. ISBN 0-7506-4171-1.
reported not to be changed by age. In contrast the time to the max-
Glass PSA. Intravenous infusion techniques: How to do it and
imal haemodynamic effect can be delayed.
why we should do it. Canadian Journal of Anaesthesia
Similarly for remifentanil the Cp50 is reduced with age, how- 1998;45:R117–127.
ever the t1/2 ke0 is prolonged, reflecting that equilibration between
Irwin MG et al. Anaesthesia 2002;57:242–248.
plasma and effect site is slower in the elderly.
Schnider TW et al. Anesthesiology 1999;90:1502–1516.
These properties suggest that induction in elderly patients
should be achieved with lower plasma concentrations than in TIVA trainer pharmacokinetic computer programme is available
from www.eurosiva.org (demo version is free).
younger adults, however it should also be titrated more slowly to
avoid side effects.

834 Bulletin 17 The Royal College of Anaesthetists January 2003