Predicted effect compartment concentration of thiopental at loss of eyelash reflex

T. A. Lim1 and K. Inbasegaran2

1Anaesthesiology Unit, Faculty of Medicine and Health Sciences, Universiti Putra Malaysia. 2Department of Anaesthesiology and Intensive Care, Hospital Kuala Lumpur, Malaysia*Corresponding author: Anaesthesiology Unit, Universiti Putra Malaysia, Aras 8, Grand Seasons Avenue, Jalan Pahang, Kuala Lumpur, Malaysia

Accepted for publication: October 31, 2000

Abstract

We derived the predicted effect compartment concentration of thiopental, at loss of the eyelash reflex, following three different injection regimens. Sixty patients were given thiopental for induction of anaesthesia. Twenty patients received multiple small boluses, 20 patients received a single bolus and 20 patients received an infusion. Computer simulation was then used to derive the effect compartment concentration. The median concentration was not significantly different between the three groups. EC50, derived after combining all three groups was 11.3 µg ml–1. The EC05–EC95 range was 6.9–18.3 µg ml–1, suggesting wide inter-individual variation.

Br J Anaesth 2001; 86: 422–4

Keywords: anaesthetics i.v., thiopental; pharmacokinetics, thiopental

The concentration at the effect compartment has a hysteresis-free relationship with the pharmacological effect. The effect compartment can be viewed as another compartment receiving and returning drug to the central compartment, but is so small it has no influence on drug pharmacokinetics. As the effect compartment concentration cannot be measured directly, determination of its concentration requires simulation of the concentration–time profile using pharmacokinetic models.

It is usually assumed that a drug obeys a single set of pharmacokinetic parameters, regardless of whether it has being given by a bolus injection or by infusion. The aim of this study is to determine the predicted effect compartment concentration of thiopental, at loss of the eyelash reflex, when given using three different injection regimens. Knowledge of this concentration will allow the planning of dosage and infusion regimens to achieve a specific pharmacodynamic end-point while minimizing the risk of over dosage.

Methods and results

The study was approved by the local clinical research ethics committee. Sixty patients, American Society of Anesthesiologists physical class I or II, scheduled for elective surgical operations gave informed consent for the study. Patients with a history of cardiovascular disease were excluded from the study. No premedication was given. On arrival in the operating theatre, an intravenous cannula was inserted into a forearm vein for infusion of drugs and fluid.

The method of induction of anaesthesia differed according to which group the patient was allocated to by randomization.

Group 1: Intermittent bolus doses of thiopental; 50 mg boluses were given every 15 s until loss of the eyelash reflex was demonstrated.

Group 2: Single bolus dose of thiopental; 4 mg kg–1 was injected over 10 s.

Group 3: Continuous infusion of thiopental at 100 mg min–1.

The eyelash reflex was tested every 2.5 s, and the time at which the reflex was lost was recorded. After induction of anaesthesia was successfully achieved, anaesthesia was maintained using a standard technique.

Effect compartment concentrations were predicted using the model reported by Stanski and Maitre.1 Serum thiopental concentrations were first calculated using standard mathematical equations which were taken from reference 2 (Hull CJ. Pharmacokinetics for Anaesthesia. Oxford: Butterworth–Heinemann, 1991) (see Appendix). These equations are incorporated into an Excel spreadsheet (V 5.0, Microsoft, Seattle, WA).2 Effect compartment concentrations were then calculated using Euler’s numerical solution.3

A sigmoid Emax model was then fitted to the concentration – effect data using the equation below:


where C is the predicted effect compartment thiopental concentration, EC50 is the predicted effect compartment concentration at which 50% of the patients had loss of the eyelash reflex, and s is a dimensionless parameter characterizing the slope of the curve of the concentration– effect relationship. The curve was fitted by unweighted least-squares non-linear regression analysis using the program Microsoft Excel. The concentrations at which 95% and 5% of patients were non-responsive (EC95 and EC05 respectively) were also calculated.

Differences between median concentrations were tested using the Kruskall–Wallis test. A value of P<0.05 was considered significant.

The mean age and weight were similar between groups (Table 1). The median effect compartment concentration, at loss of the eyelash reflex, was not significantly different between groups (Table 1). In contrast, there was a significant difference in the median serum concentration (P<0.05). EC50, derived after combining data from all three groups was 11.3 µg ml–1, and s was 6.03. The EC05–EC95 range was 6.9–18.3 µg ml–1.


View this table:
[in this window]
[in a new window]
 
Table 1 Patient data (mean (SD)), predicted effect compartment and serum concentrations (median (range)). Median effect compartment concentrations were not significantly different between groups. Predicted median serum concentrations were significantly different (P<0.05)
 
Comment

The median effect compartment concentration reported in this study is similar to those reported by previous investigators using loss of verbal response or loss of voluntary motor power as the end-point.4 5 The median effect compartment concentration we derived is also similar to that obtained by computer simulation for jugular venous concentration at awakening from anaesthesia.6 The jugular venous concentration reflects the brain concentration, and like the effect compartment concentration should be hysteresis-free. This would explain why the concentration and loss of consciousness and awakening are similar.

The median concentration for Group 3 (infusion) appears to be higher than that for Group 1 or 2, although this did not reach statistical significance. One reason for this could be the difficulty in accurately defining the point at which loss of the eyelash reflex occurred in the infusion group. In Groups 1 and 2, the effect compartment concentration at induction of anaesthesia was increasing rapidly because of the high serum–effect compartment concentration difference. In Group 3, the increase in effect compartment concentration was much slower, leading to a more gradual onset of effect. This made the onset of anaesthesia less distinct and could have led to a delay in the recording of the end-point.

In our study, the predicted effect compartment concentration at the pharmacodynamic end-point did not correlate with the age or weight of the patient. This could be because the pharmacokinetic-pharmacodynamic model incorporated both age and weight as covariates. This is an advantage as it avoids over dosage in elderly and thin patients.

Most manual dosing regimens and target controlled infusion systems rely on a series of bolus injections and infusion rates to achieve a desired plasma or effect compartment concentration. Optimal use of such infusion systems requires some form of real time estimation of drug concentration. Differences in the disposition of drugs cause the actual concentration achieved after a specific infusion regimen to differ from one patient to another. Knowing the measured concentration will allow an accurate correlation with effect. However, technology for such measurements is unlikely to be available in the near future. Real time prediction of the drug concentration offers a reasonable alternative at the present time.

Pharmacodynamic effect is assumed to be proportional to the concentration of the drug in the region of the receptors. Ideal infusion systems should be able to predict this concentration reliably despite the dosing history. The effect compartment concentration may not be the true concentration around the receptor sites, but this does not matter so long as it can be correlated with effect.

Target controlled infusion devices currently in use incorporate algorithms to predict drug concentrations. Results from our study suggest that concentrations predicted in such a fashion can be correlated to effect. While thiopental is not used for maintenance of anaesthesia, target controlled infusions of thiopental can still be used for induction. It is likely that other intravenous hypnotics will behave in a similar fashion, although this will need to be investigated further.

In conclusion, we found the predicted effect compartment concentration of thiopental at loss of the eyelash reflex to be independent of the injection regimen.

Appendix

Calculation of the predicted serum and effect compartment concentrations
To calculate the hybrid rate constants {alpha}, ß, {delta}2


View this table:
[in this window]
[in a new window]
 
Table 2 Pharmacokinetic model
 
Let:


To calculate the instantaneous change in central compartment concentration with a bolus dose

C{phi}1 = C{phi}1 + (Xd / V1)

To calculate change in central and peripheral compartment concentrations during an infusion and/or after a bolus dose

F1 = (k'01 / V1) + C{phi}1(k21 + k31) + C{phi}2k12 + C{phi}3k13

G1 = (k'01 / V1)(k21 + k31) + C{phi}1k21k31 + C{phi}2k12k31 + C{phi}3k13k21

H1 = (k'01 / V1)k21k31

J1 = C{phi}1

F2 = C{phi}1k21 + C{phi}2(k12 + k13 + k31 + k10)

G2 = (k'01 / V1)k21 + C{phi}1k21k31 + C{phi}2k31(k12 + k10) + C{phi}3k13k21

H2 = (k'01 / V1)k21k31

J2 = C{phi}2

F3 = C{phi}1k31 + C{phi}3(k12 + k13 + k21 + k10)

G3 = (k'01 / V1)k31 + C{phi}1k21k31 + C{phi}2k31k12 + C{phi}3k21(k13 + k10)

H3 = (k'01 / V1)k21k31

J3 = C{phi}3

Ai = [{alpha}(Ji{alpha}Fi) + Gi – (Hi / {alpha})] / [(ß {alpha})({delta}{alpha})]

Bi = [ß(Jiß – Fi) + Gi – (Hi / ß)] / [({alpha} – ß)({delta} – ß)]

Di = [{delta}(Ji{delta}Fi) + Gi – (Hi / {delta})] / [({alpha}{delta})(ß {delta})]

E = k'01 / (V1k10)

Finally,

Ci = Aie{alpha}t + Bie–ßt + Die{delta}t + E

where C1, C2 and C3 are the central, shallow peripheral and deep peripheral concentrations respectively.

To calculate the change in effect compartment concentration using Euler’s numerical solution

The volume of the effect site (V4) was arbitrarily defined as V1 / 10000

k41 = keo

k14 = keo / 10000

To calculate the change in effect compartment concentration at each time interval:

{Delta}A4 = [ k14A1k41A4 ]. {Delta}t

where A1 and A4 are the mass of drug in the central and effect compartments respectively, and {Delta}t is the time interval. The time interval used in the simulation was 2.5 s. Therefore at any time, t:

A4 (t) = A4 (t–2.5) + {Delta}A4

C4 (t) = A4 (t) / V4

A1 (t) = C1 (t) x V1

where C1 and C4 are the concentrations in the central compartment and effect compartments respectively.

References

1 Stanski DR, Maitre PO. Population pharmacokinetics and pharmacodynamics of thiopental: the effect of age revisited. Anesthesiology 1990; 72: 412–22[ISI][Medline]

2 Hull CJ. Pharmacokinetics for Anaesthesia. Oxford: Butterworth-Heinemann, 1991

3 Maitre PO, Shafer SL. A simple pocket calculator approach to predict anesthetic drug concentrations from pharmacokinetic data. Anesthesiology 1990; 73: 332–6[ISI][Medline]

4 Hung OR, Varvel JR, Shafer SL, Stanski DR. Thiopental pharmacodynamics. II. Quantitation of clinical and electroencephalographic depth of anesthesia. Anesthesiology 1994; 80: 1216–27[ISI][Medline]

5 Shanks CA, Avram MJ, Krejcie TC, Henthorn TK, Gentry WB. A pharmacokinetic-pharmacodynamic model for quantal responses with thiopental. J Pharmacokinet Biopharm 1993; 21: 309–21[ISI][Medline]

6 Barratt RL, Graham GG, Torda TA. Kinetics of thiopental in relation to site of sampling. Br J Anaesth 1984; 56: 1385–90[Abstract]