Biphasic EEG changes in relation to loss of consciousness during induction with thiopental, propofol, etomidate, midazolam or sevoflurane

K. Kuizenga1, J. M. K. H. Wierda1 and C. J. Kalkman2

1Department of Anesthesiology, University Hospital of Groningen, PO Box 30.001, 9700 RB Groningen, The Netherlands. 2Department of Anesthesiology, University Medical Centre Utrecht, Heidelberglaan 100, PO Box 85500, 3508 GA Utrecht, The Netherlands*Corresponding author

Accepted for publication: November 9, 2000


    Abstract
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
The time course of four EEG effect variables, amplitude in the 2–5 Hz and in the 11–15 Hz band, spectral edge frequency 95% (SEF95), and bispectral index (BIS), in response to increasing concentrations of thiopental, propofol, etomidate, midazolam, or sevoflurane during a 10 min induction of anaesthesia was studied in 25 patients to determine the existence of a biphasic effect and to study the relationship of the EEG effect to the moment of loss of consciousness. A biphasic effect, that is, an initial increase of the effect variable followed by a decrease at higher concentrations, during the transition from consciousness to unconsciousness was found in EEG amplitude (both frequency bands) and in SEF95 for all anaesthetics except midazolam. There was a concentration-related decrease in BIS for all anaesthetics. There was no consistent relationship between the time of occurrence of the peak EEG effect, or the value of the EEG variable and the moment of loss of consciousness. With rapidly changing drug concentrations during the induction of anaesthesia, none of these EEG effect variables could be correlated to the moment of loss of consciousness.

Br J Anaesth 2001; 86: 354–60

Keywords: brain, electroencephalography; anaesthesia, depth; anaesthetics i.v., thiopental; anaesthetics i.v., propofol; anaesthetics i.v., etomidate; hypnotics benzodiazepine, midazolam; anaesthetics volatile, sevoflurane


    Introduction
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Processed EEG variables recorded during anaesthesia usually decrease with increasing drug concentrations. However, during the transition from consciousness to unconsciousness, using induction agents, many derived EEG variables show biphasic effects, that is an initial increase of the effect variable followed by a decrease at higher concentrations. This has been observed with thiopental,1 propofol,2 3 and sevoflurane.4 For etomidate, a biphasic effect was described in animals.5 Midazolam does not appear to induce biphasic EEG changes.6 The moment of loss of consciousness seems to be related to the occurrence of a maximum effect of the EEG variable, at least when propofol, thiopental, or sevoflurane are used to induce anaesthesia. For the other drugs such a relationship has not been demonstrated. If a biphasic relationship exists for all hypnotic agents, then the reappearance of a maximum EEG height might be used as an indicator of imminent awareness.

We therefore studied the behaviour of four different EEG effect variables during slow induction of anaesthesia with thiopental, propofol, etomidate, midazolam, or sevoflurane with the aims of determining whether a biphasic EEG response was present, and to assess the time to the maximum EEG effect, the times of loss of consciousness, and the relationship between these variables for each study drug. The following variables were analysed: EEG amplitude in two frequency bands, spectral edge frequency 95%, and the bispectral index (BIS), a proprietary composite EEG effect variable designed to have no biphasic response pattern.


    Patients and methods
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
After institutional board approval and obtaining written informed consent, 25 patients, ASA grade I and II undergoing elective vertebral disk surgery entered the study. Patients with a history of recent intake of drugs that might affect the EEG, an alcohol intake in excess of 30 g day–1, neurological disorders, extreme nervousness, or a contraindication for one of the study drugs were excluded. No oral intake was allowed after midnight preceding the day of surgery. No pre-medication was administered.

After arrival in the anaesthetic room, an i.v. cannula was inserted and 0.9% saline 500 ml was infused rapidly for fluid loading. Patients were monitored with a three-lead ECG, SpO2, and non-invasive arterial pressure measurements performed at 1-min intervals.

Each experiment took place before the start of surgery and lasted 15 min. The study period was divided into a 3 min baseline EEG recording, a 10 min duration EEG recording during study drug administration, and was followed by a 2 min EEG recording without drug administration. The calculation of the dose body weight, corrected for height, was: height (cm)–100=kg body weight. The patients were randomly assigned to one of five groups to receive a 10 min constant rate infusion of thiopental 1 mg kg–1 min–1, propofol 0.5 mg kg–1 min–1, etomidate 0.06 mg kg–1 min–1, midazolam 0.03 mg kg–1 min–1, or sevoflurane in oxygen enriched air administered by a circle system with a fresh gas flow of 9 litre min–1 and facemask. The sevoflurane concentration of the fresh gas was initially set at 2% and increased every minute by an additional 2% to a maximum of 8%. The last concentration was maintained for a further 7 min. With this dosing regimen we aimed to induce a steadily increasing blood and CNS concentration (Fig. 1).



View larger version (13K):
[in this window]
[in a new window]
 
Fig 1 Time course of expected plasma concentrations of thiopentone, propofol, etomidate, and midazolam based on population kinetics (Stanpump) and of sevoflurane concentrations in the fresh gas flow.

 
If spontaneous ventilation became insufficient as indicated by SpO2 below 92% or signs of upper airway obstruction, a free airway was re-instituted by chin lift and if necessary ventilation was assisted by facemask and oxygen enriched air (40%) to maintain E'CO2 between 4.5 and 6.0 kPa.

EEG was recorded from Cz-Fpz and the M2 (right mastoid)-Fpz2 with an electrode placed between Fpz and Fpz2 as reference. All signals were recorded with the patients in the supine position and eyes closed. The analogue EEG was recorded using the pre-amplifiers of the Lifescan EEG monitor and stored on tape for off-line analysis. For calculation of EEG amplitude in the 0–5 and 11–20 Hz band we used aperiodic analysis7 (Lifescan, software version 4.3, Diatek, San Diego, CA, USA). For calculation of BIS8 and spectral edge frequency 95% (SEF95) values we used the Aspect A-1000 EEG monitor (BIS version 3.12: Aspect Medical Systems Inc., Natick, MA, USA). EEG amplitudes in the two frequency bands were calculated from 10-s epochs. Epochs containing an EEG amplitude greater than 30 µV in the 25–30 Hz band were considered to contain too many artefacts and were rejected. The Aspect A1000 low pass filter was set at 30 Hz, the high pass filter was set to 2 Hz. Periods containing artefacts, as indicated by the Aspect A1000 were rejected. Spectral edge smoothing was set at 10 s and BIS smoothing at 15 s. No correction was made for the internal 60 s smoothing of the BIS 3.12. Thus, amplitude and SEF95 data represent EEG data of the preceding 10 s. BIS data represent data from the preceding 75 s.

Responsiveness was determined by testing the response of the patient to simple commands from a pre-recorded tape (‘raise your thumb’, ‘spread your fingers’, and ‘clench your fist’), given by headphones every 30 s. The first time that the patient did not respond to a verbal command was registered as the time of loss of responsiveness (LR). The times of LR were related to the values of the EEG derived variables recorded at that moment.

Definitions and statistical analysis
The EEG response to increasing blood concentrations was considered biphasic if after the start of the drug administration the EEG effect variable increased at least 2 SD above baseline value and, thereafter, decreased 20% below that maximum value.

The patient was supposed to have lost consciousness at the moment of the maximum change of the EEG variable if the moment of LR was less than 30 s different from it.

Data are presented as median and range. EEG data were compared using the Kruskal-Wallis test. P<0.05 was considered significant.


    Results
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Physical characteristics of the patients and the moments of LR are presented in Table 1. EEG data from the Fpz-Cz leads during the infusion showed a similar time course compared with the EEG data from the Fpz2-M2 leads. But EEG data derived from the Fpz2-M2 leads showed more variability than EEG data derived from the FPz-Cz leads during the baseline recording before drug administration. We therefore decided to present only data from the Fpz-Cz leads.


View this table:
[in this window]
[in a new window]
 
Table 1 Physical characteristics of the patients and the moment of loss of consciousness after the start of the study. Data are expressed as median (range)
 
During the administration of thiopental, propofol, etomidate, and sevoflurane the EEG of the patients showed an increase of alpha and beta activity followed by an increase of delta activity and a decrease of beta and alpha activity. All patients in the propofol and etomidate groups and most of the patients in the thiopental group showed a decrease in delta activity by end of the infusion. In the sevoflurane patients, delta activity initially decreased after the maximal effect but did not decrease further despite continued administration. Increasing midazolam doses resulted in increased beta activity.

The EEG amplitude in both frequency bands and SEF95 showed biphasic effects for all drugs except midazolam. The BIS value decreased for all drugs (Table 2, Figs 2 and 3).


View this table:
[in this window]
[in a new window]
 
Table 2 Occurrence of biphasic response (n); the number of patients losing consciousness (n) before and after the peak EEG effect (when present); the median time (min) between peak EEG effect and loss of consciousness; and the median EEG value (range) at the moment of loss of consciousness of the EEG amplitude (µV) in the 2–5 Hz band, the EEG amplitude (µV) in the 11–20 Hz band. Spectral edge frequency 95% (Hz) and BIS. *Different from other drugs (P<0.05)
 


View larger version (33K):
[in this window]
[in a new window]
 
Fig 2 EEG amplitude versus time in the 2–5 and 11–20 Hz bands for individual patients. (Open circles indicate the moment of loss of consciousness.)

 


View larger version (32K):
[in this window]
[in a new window]
 
Fig 3 Spectral edge frequency 95% and BIS versus time for individual patients. (Open circles indicate the moment of loss of consciousness.)

 
We did not observe a consistent relationship between the moment of LR and the moment of occurrence of a peak EEG effect. The value of the EEG effect at the moment of LR showed a large variability between patients. BIS values at the moment of LR were significantly higher in the sevoflurane group than in all other groups (Table 2).


    Discussion
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
In this study we demonstrated biphasic EEG effects for all the induction agents except for midazolam. Biphasic effects, that is an increase in alpha and beta activity followed by a decrease in alpha and beta activity and a simultaneous increase in delta activity have been observed in the analogue EEG.9 10 Veselis and co-workers have shown that EEG power in the beta band increases with increasing blood concentrations of thiopental, propofol, and midazolam.11 They did not observe a decrease of EEG power, but as the drugs were administered in sedative doses only, their patients did not lose consciousness. Other investigators have shown that SEF95 and median frequency decrease when depth of anaesthesia increases.1214 Therefore, a peak change in EEG amplitude in the higher frequency bands during the transition from sedation to unconsciousness is very likely. Such biphasic effects have been described for thiopental,15 propofol,16 and more recently for sevoflurane.4 The occurrence of a short-lived peak effect on the EEG might be a marker for the transitional state between consciousness and unconsciousness. If consciousness is lost before the peak change in EEG occurs, that is at a lower blood concentration, during decreasing concentrations a peak EEG effect is coupled with a risk of imminent awareness. Although we observed biphasic effects for all drugs except midazolam during increasing drug concentrations, loss of consciousness did not occur consistently before the maximum change in these EEG variables. This finding limits the potential application of the occurrence of a peak EEG effect to predict imminent awareness. The actual EEG amplitudes and SEF95 values at which consciousness was lost showed a large inter-individual variability. Therefore, absolute values of EEG derived variables are not reliable indicators of loss of consciousness either.

One explanation for the absence of a consistent correlation between the time to loss of consciousness and the value of the EEG effect variable or the time to peak EEG effect for the five drugs might be the biphasic response of the EEG variables to increasing blood concentrations. Such a biphasic response may be the result of simultaneous drug effects both on systems that cause EEG activation and on systems that cause EEG suppression. At lower concentrations, activation is more pronounced than suppression, which results in an increase of amplitude at higher frequencies. At higher concentrations the suppression becomes more evident which results in a decrease of amplitude at higher frequencies and eventually in a decrease of EEG amplitude at lower frequencies. All five drugs suppress consciousness in a dose dependent way. Midazolam does not induce a biphasic EEG response and consciousness was lost before maximal beta activity was reached. With the other drugs one might expect that consciousness would be lost when EEG activity was still increasing in a similar way to midazolam. However, EEG suppression at low concentrations might have counteracted the activation to such an extent that EEG activity had already started to decrease before consciousness was lost. As all drugs have different chemical structures it is likely that there will be differences in the balance of effects on activation and suppression. Therefore, such differences will result in differences in EEG effects at the moment of loss of consciousness.

BIS has been used successfully as an indicator of the level of sedation and hypnosis.17 18 In the present study we observed that consciousness in the sevoflurane group was lost at higher BIS values than in the other groups (Table 2). This might be caused by the steep decrease in BIS during induction. As a result of the long period over which BIS 3.12 calculates its value, for example, 60+15 s smoothing, the actual BIS value will be overestimated if BIS decreases rapidly. More recent software versions with a shorter calculation epoch may overcome this problem. Another explanation for the higher BIS values observed at the time of LR might be because of the non-specific mechanism of action of sevoflurane inducing loss of consciousness at lower levels of cortical suppression. In a study in which semi-steady state concentrations of sevoflurane were applied,4 a method that will effectively eliminate averaging effects, the average BIS for preventing response to verbal command was 73. This value is similar to the values found for midazolam, thiopental, propofol, and etomidate in our study. This finding makes the first explanation for the higher BIS values at loss of responsiveness for sevoflurane more likely.

A drawback of the present study is the lack of EEG data during return of consciousness. EEG values at return of consciousness might be different from the EEG values at loss of consciousness and variability might be less. More gradual concentration changes during emergence from anaesthesia will decrease inaccuracy in determination of EEG effect and the determination of the moment of return of consciousness will be less critical. However, other investigators19 have already indicated that during repeated transitions from unconsciousness to consciousness as the result of repeated interruptions of propofol administration to patients under spinal analgesia, neither SEF95 nor BIS are accurate indicators of imminent awareness.

We conclude that thiopental, propofol, etomidate, and sevoflurane, but not midazolam induce biphasic EEG effects during the transition from consciousness to unconsciousness. There is no consistent time relationship between the peak EEG effect and the moment of LR. In the BIS algorithm, the biphasic response has been linearized effectively, resulting in a progressive decrease of BIS during progressively increasing hypnotic drug concentrations. Both the large inter-subject variability in the values of EEG amplitude and spectral edge, and the relatively long averaging period of BIS, limit the applicability of these variables as markers for imminent consciousness.


    Acknowledgements
 
We thank Steven L.Shafer for the use of the STANPUMP software, which is freely available from the author, Steven L.Shafer, MD, Anesthesiology Service (112A), PAVAMC, 3801 Miranda Ave., Palo Alto, CA 94304, USA.


    References
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
1 Hung OR, Varvel JR, Shafer SL, Stanski DR. Thiopental pharmacodynamics. II. Quantitation of clinical and electroencephalographic depth of anesthesia. Anesthesiology 1992; 77: 237–44[ISI][Medline]

2 Billard V, Gambus PL, Chamoun N, Stanski DR, Shafer SL. A comparison of spectral edge, delta power, and bispectral index as EEG measures of alfentanil, propofol, and midazolam drug effect. Clin Pharmacol Ther 1997; 61: 45–58[ISI][Medline]

3 Kuizenga K, Kalkman CJ, Hennis PJ. Quantitative electroencephalographic analysis of the biphasic concentration-effect relationship of propofol in surgical patients during epidural analgesia. Br J Anaesth 1998; 80: 725–32[Abstract/Free Full Text]

4 Katoh T, Suzuki A, Ikeda K. Electroencephalographic derivatives as a tool for predicting the depth of sedation and anesthesia induced by sevoflurane. Anesthesiology 1998; 88: 642–50[ISI][Medline]

5 De Paepe P, Belpaire FM, Van Hoey G, Boon PA, Buylaert WA. Influence of hypovolemia on the pharmacokinetics and the electroencephalographic effect of etomidate in the rat. J Pharmacol Exp Ther 1999; 290: 1048–53[Abstract/Free Full Text]

6 Buhrer M, Maitre PO, Hung O, Stanski DR. Electroencephalographic effects of benzodiazepines. I. Choosing an electroencephalographic parameter to measure the effect of midazolam on the central nervous system. Clin Pharmacol Ther 1990; 48: 544–54[ISI][Medline]

7 Gregory TK, Pettus DC. An electroencephalographic processing algorithm specifically intended for analysis of cerebral electrical activity. J Clin Monit 1986; 2: 190–7[Medline]

8 Sigl JC, Chamoun NG. An introduction to bispectral analysis for the electroencephalogram. J Clin Monit 1994; 10: 392–404[ISI][Medline]

9 Clark DL, Rosner BS. Neurophysiologic effects of general anesthetics. I. The electroencephalogram and sensory evoked responses in man. Anesthesiology 1973; 38: 564–82[ISI][Medline]

10 Hazeaux C, Tisserant D, Vespignani H, Hummer S, Kwan N, Laxenaire MC. [Electroencephalographic impact of propofol anesthesia] Retentissement électroencéphalographique de l’anésthesie au propofol. Ann Fr Anesth Reanim 1987; 6: 261–6[Medline]

11 Veselis RA, Reinsel R, Sommer S, Carlon G. Use of neural network analysis to classify electroencephalographic patterns against depth of midazolam sedation in intensive care unit patients. J Clin Monit 1991; 7: 259–67[ISI][Medline]

12 Schwilden H, Stoeckel H, Schuttler J. Closed-loop feedback control of propofol anaesthesia by quantitative EEG analysis in humans. Br J Anaesth 1989; 62: 290–6[Abstract]

13 Stanski DR, Hudson RJ, Homer TD, Saidman LJ, Meathe E. Pharmacodynamic modeling of thiopental anesthesia. J Pharmacokinet Biopharm 1984; 12: 223–40[ISI][Medline]

14 Schwender D, Daunderer M, Klasing S, Finsterer U, Peter K. Power spectral analysis of the electroencephalogram during increasing end-expiratory concentrations of isoflurane, desflurane and sevoflurane. Anaesthesia 1998; 53: 335–42[ISI][Medline]

15 Buhrer M, Maitre PO, Hung OR, Ebling WF, Shafer SL, Stanski DR. Thiopental pharmacodynamics. I. Defining the pseudo-steady-state serum concentration-EEG effect relationship. Anesthesiology 1992; 77: 226–36[ISI][Medline]

16 Dutta S, Matsumoto Y, Gothgen NU, Ebling WF. Concentration-EEG effect relationship of propofol in rats. J Pharm Sci 1997; 86: 37–43[ISI][Medline]

17 Flaishon R, Windsor A, Sigl J, Sebel PS. Recovery of consciousness after thiopental or propofol. Bispectral index and isolated forearm technique. Anesthesiology 1997; 86: 613–9[ISI][Medline]

18 Glass PS, Bloom M, Kearse L, Rosow C, Sebel P, Manberg P. Bispectral analysis measures sedation and memory effects of propofol, midazolam, isoflurane, and alfentanil in healthy volunteers. Anesthesiology 1997; 86: 836–47[ISI][Medline]

19 Gajraj RJ, Doi M, Mantzaridis H, Kenny GN. Analysis of the EEG bispectrum, auditory evoked potentials and the EEG power spectrum during repeated transitions from consciousness to unconsciousness. Br J Anaesth 1998; 80: 46–52[ISI][Medline]