Secondary Logo

Journal Logo

Update in Intravenous Anaesthesia: Original Papers

How to manage drug interactions

Glass, P. S. A.; Howell, S.; Gan, T. J.; Ginsberg, B.

Author Information
European Journal of Anaesthesiology: May 1997 - Volume 14 - Issue - p 33-39



Anaesthesia consists of hypnosis, analgesia and amnesia. As yet there is no single intravenous anaesthetic drug that can effectively and safely provide all of these three components of anaesthesia. Thus, to provide the anaesthetic state, usually two or more drugs are combined. It is well known that one drug may readily alter the disposition (pharmacokinetics) of a second drug. In addition, as the anaesthetic state may be produced by drugs acting at a variety of receptor sites, it is not unexpected that their resultant combined effect (pharmacodynamics) will produce complex interactions. The interaction between two volatile anaesthetics has been shown to be simply additive; i.e. their combined effect is the result of adding their individual effects [1]. However, because of both the pharmacokinetic and pharmacodynamic interactions, when intravenous anaesthetic drugs are combined, this simple additive effect is unusual and a more complex interaction is normally observed. Thus, it is important to be aware of the drug interactions that occur when combining intravenous anaesthetics.

Pharmacodynamic interactions

A pharmacodynamic interaction implies a change in the observed effect when one drug is combined with another, compared to their effect when given alone, and this change in the observed effect is not a result of changes in the drugs' concentration in the biophase (effect) site. Thus, pharmacodynamic interactions exclude interactions occurring as a result of one drug affecting the pharmacokinetics of the second drug. Pharmacodynamic interactions occur as a result of several mechanisms, most of which are presently ill understood [2,3]. At the cellular level, one drug may enhance the binding of a second drug to its receptor or conversely inhibit its binding (e.g. agonist antagonist). A drug may also alter the intracellular signal transduction pathway of another drug (e.g. the potentiation of the arrhythmogenic effects of β-agonists by volatile anaesthetics by both increasing adenyl cyclase activity, or the increased MAC in alcoholics due to development of tolerance of the GABAergic receptor), or one drug may effect the uptake or production of neurotransmitters whose release are altered by the second drug (e.g. reversal of neuromuscular blockers by anticholinesterases). A pharmacodynamic interaction may also occur as a result of two drugs acting on two separate receptor systems but whose final common pathway either at the cellular or subcellular levels are similar. This latter mechanism is probably the most common cause for the pharmacodynamic drug interactions seen between intravenous drugs used to provide anaesthesia. With the cloning of many of the receptors involved in anaesthesia, we are obtaining a greater understanding of drug receptor binding, receptor signal transduction, and the resultant pharmacological effect. With this should also come a greater understanding of pharmacodynamic drug interactions and the possibility of being able to predict the likely interactions that will occur with the introduction of newer drugs.

There are several ways of trying to establish the interaction of drugs in providing anaesthesia. In addition, there are different endpoints (i.e. measures of effect) that may be used to assess these interactions. The two most commonly used measures of assessing the adequacy of anaesthetic effect are loss of response to a verbal command and purposeful movement to a skin incision. These two measures of effect clearly measure different endpoints, and thus the interaction of the same two drugs for each of these endpoints may be different. Both of these endpoints have clinical significance and guide our dosing regimens for loss of consciousness; i.e. induction and for an adequate anaesthetic state during surgery.

Kissin and co-workers were the first to try and systematically establish the interaction of various intravenous opioids and hypnotics in providing loss of consciousness [4-18]. The drug interactions were assessed by the use of isobolograms whereby a bolus dose response for each drug for loss of consciousness is obtained. A line drawn between the ED50 values for each of the individual drugs represents the line where the combination of the two drugs would also result in a 50% loss of consciousness if the interaction between the two drugs were simply additive (Fig. 1). The combinations of the two drugs that would be estimated from this line to produce loss of consciousness in 50% of patients is then tested. If the combination results in what is predicted from the line joining the ED50s, then the interaction is simply additive. If the resultant combination is below the line of additivity (i.e., smaller doses than predicted are required to produce a 50% response), then the drugs may be considered to be acting synergistically. If the interaction lies above the line of additivity then the interaction is considered to be antagonistic. The vast majority of the interactions of the intravenous anaesthetics tested have shown a synergistic interaction, although some combinations have shown simple additive and even antagonistic [6,12] interactions. Also, the degree of the interactions are not consistent across the same class of drugs [4] nor across the different endpoints of anaesthesia (e.g. the same degree of interaction is not true for both loss of consciousness and response to skin incision with the same two drugs) [19]. The degree of synergism is also not consistent over the entire dosage range; i.e. 10% of the ED50 of drug A and 20% of the ED50 of drug B may provide an ED50, but that does not mean that this will also occur with a 20% ED50 dose of drug A and a 10% ED50 dose of drug B. Thus each specific dose combination results in a specific result independent of the results from other combinations. This makes the results from these data of limited clinical use other than to make the clinician aware that significant synergism may exist and that they must be careful in dosing when combining intravenous anaesthetics. In general, the combination of a benzodiazepine with opioids or any other hypnotic results in a marked synergistic interaction [9-13, 15-22]. The interaction between opiates and other hypnotics is also synergistic [4,7,19,23], but not as marked as with benzodiazepines. Morphine appears to provide a greater synergism with hypnotics than the piperidine opiates. The interaction between three drugs for loss of consciousness has also been studied. It appears that the addition of the third drug does not necessarily provide a further synergistic interaction [19,24].

Fig. 1
Fig. 1:
An isobologram of the interaction of thiopentone and morphine to induce loss of consciousness in rats. A, E = ED50 values for thiopentone and morphine given alone. B, C, D = ED50 values for thiopentone-morphine combinations. The dashed straight line is an additive line. The solid lines connecting adjacent ED50 points represent the ED50 isobol. RR = righting reflex. Reproduced with permission from [7].

When using isobolograms to assess the interaction between two drugs there is no attempt at discerning if the synergistic interaction is due to an effect on the drugs' pharmacokinetics resulting in higher concentrations, or due to a pharmacodynamic effect; i.e. a greater sensitivity to the combination. In addition, all drugs demonstrate some delay between dosing and peak effect [25]. This delay is different for each drug and thus it is possible that many of the interactions are not measured at the peak effect of both drugs, thereby providing a false impression of their interaction. A different approach to evaluating the interaction between the intravenous anaesthetics is to administer both drugs to a constant target concentration and only assess their effect once the plasma concentration and biophase (effect compartment) have equilibrated. With the advent of pharmacokinetic model-driven drug delivery systems, intravenous anaesthetics can be administered to target concentrations (much like volatile anaesthetics), and this concentration maintained until equilibration between plasma and the biophase has occurred [25]. In these types of studies, any pharmacokinetic effects of the drug combinations are excluded and only the pharmacodynamic interaction is evaluated.

Probably the most commonly used combination of anaesthetics is isoflurane and fentanyl. Using computer-assisted continuous infusion (CACI) to maintain constant fentanyl concentration, McEwan et al.[26] have recently shown that the MAC of isoflurane is markedly reduced initially, with a 50% MAC reduction produced by 1.7 ng ml−1 fentanyl (Fig. 2). However, beyond 3 ng ml−1, a plateau or ceiling effect is seen with a maximum MAC reduction of approximately 80%. Thus, the maximum reduction in isoflurane was to a concentration of ± 0.3%, close to the MACawake for isoflurane. In addition, the minimum effective analgesic concentration of fentanyl is 0.6 ng ml−1, so that the steepest reduction in MAC occurs within the analgesic concentration range of fentanyl (i.e. 1-2 ng ml−1). Clinically, significant respiratory depression also occurs with plasma fentanyl concentrations greater than 2 ng ml−1.

Fig. 2
Fig. 2:
The interaction between fentanyl and isoflurane. The solid line represents the concentration of the two drugs that will prevent movement at skin incision in 50% of patients.

Alfentanil [27], sufentanil [28], and remifentanil [29] produce similar reductions in isoflurane MAC, with an initial steep reduction at lower concentrations and a plateau effect at higher concentrations. The concentration producing a 50% reduction of the MAC of isoflurane provides a means of determining equipotency (in the concentration domain) between the opiates thus far tested. Remifentanil has an extremely short context-sensitive half-time (3-5 min) [30-32]. The MAC reduction of isoflurane by remifentanil has recently been published [29]. The results of this drug interaction are identical to those with the other opioids thus far tested. A 50% MAC reduction was obtained with a remifentanil concentration of 1.4 ng ml−1. Thus, in the concentration domain, remifentanil is almost equipotent with fentanyl. As remifentanil has such a brief context-sensitive decrement time, it has been possible to administer remifentanil to extremely high concentrations. Again, even with these very high concentrations, a ceiling effect was still observed with the ceiling at an isoflurane concentration of 0.2-0.3%.

To evaluate the interaction of two drugs for total intravenous anaesthesia, a similar MAC/Cp50 (plasma concentration at which 50% of patients will not show purposeful movement at skin incision), a reduction study of propofol by fentanyl was done. This study evaluated both the interaction of these two drugs for loss of response to verbal command and for the prevention of movement to skin incision [33]. Although studies with isobolograms showed a synergistic interaction between fentanyl and propofol for loss of consciousness, Smith et al.[33] could only demonstrate a moderate decrease in propofol concentrations with increasing fentanyl concentrations (up to 10 ng ml−1). Telford et al. have demonstrated a similar lack of interaction between fentanyl (1-3 ng ml−1) and thiopentone to induce loss of consciousness [34]. However, for prevention of movement at skin incision, a marked interaction was seen similar to that observed between fentanyl and isoflurane. The Cp50 of propofol alone for skin incision (arterial sampling) was 16 μg ml−1. The Cp50 was markedly reduced by low concentrations of fentanyl and again reached a plateau beyond 3 ng ml−1 fentanyl. The propofol concentration at the maximal Cp50 reduction was approximately 2.5 μg ml−1, which corresponded with the Cp50 asleep for propofol. Thus, the interaction of propofol with fentanyl appears very similar to the interaction of opioids with the volatile anaesthetics. Vuyk et al. have completed similar studies determining the interaction of alfentanil and propofol [35]. The results for absence of movement at skin incision with the combination of these two drugs was again very similar to that seen with propofol and fentanyl and the interaction with isoflurane and opioids. Vuyk et al. took this interaction one step further, in that they also observed the time to awakening at each of these combinations. Thus not only were they able to define the optimal interaction for the prevention of a response to skin incision but also the implication of these concentrations on recovery. The context-sensitive decrement times [36] provide an indication of the time to recovery. In general, the longer the duration of drug administration, the longer the recovery time will be. Also, as higher concentrations of drugs are administered, so does the percentage decrement required to achieve awakening increase. The increasing time for recovery will differ markedly between drugs. For example, if the decrement time for drug A is increased to 55%, this may imply only a 3-min delay in recovery, while for drug B a similar increase in decrement time to 55% will result in a 30-min delay in recovery time.

The differences in recovery time according to which drug is increased is well illustrated in Fig. 3, which shows that optimal recovery time occurs at an alfentanil concentration of approximately 80 ng ml−1 and propofol of approximately 3 μg ml−1. When the concentration of propofol is increased, the concentration of alfentanil can be decreased, but the overall time for recovery increases. Similarly, as the concentration of alfentanil increases, the concentration of propofol can be decreased, but the time for recovery increases. It can be seen that when the concentration of alfentanil is increased beyond 80 ng ml−1, even though the concentration of propofol can be reduced, there is a marked increase in the time for recovery. This increase in recovery time is much larger than the increase in recovery time that occurs when propofol is increased beyond 3 μg ml−1. Thus, the clinical implications of this drug interaction between propofol (or volatile anaesthetic) with either fentanyl, alfentanil or sufentanil to provide anaesthesia is that the infusion regimen should provide an analgesic concentration of the opioid equivalent to 1-2 ng ml−1 of fentanyl (see Table 1). The propofol infusion should provide an absolute minimum concentration of 2.5 μg ml−1 which corresponds to a loading dose of 1.5 mg kg−1 followed by an infusion rate of approximately 50-80 μg kg−1min−1, or when administering isoflurane, a minimum concentration of 0.4%. If the patient demonstrates signs of inadequate anaesthesia it is preferable to increase either the propofol or the volatile anaesthetic, as increasing these has less of an effect on prolonging wake-up time than increasing the opioid. Remifentanil has an extremely short context-sensitive half-time of only 3-5 min and a context-sensitive 80% decrement time of 10-15 min irrespective of the duration of the infusion. Thus, with remifentanil the reverse is true. It is now preferable to administer remifentanil to high opioid concentrations of 8-12 ng ml−1 (0.25-0.4 μg kg−1min−1) with just sufficient hypnotic to ensure an unconscious patient. Also, if the patient responds, recovery time is far less prolonged by increasing the remifentanil than by increasing propofol or isoflurane.

Fig. 3
Fig. 3:
The interaction between propofol and alfentanil plasma concentration in preventing movement at skin incision. Also illustrated is the time to recover as either the alfentanil or propofol concentrations were increased. Reproduced with permission from[35].
Table 1
Table 1:
Manual opioid infusion schemes

As already mentioned, the major effects measured in drug interaction studies have been loss of consciousness and response to skin incision. There are very few studies that have looked at the effect of these interactions on amnesia, yet this is a critical part of the anaesthetic state. In an unreported study by our group, minimal potentiation was found of the amnesic effects of midazolam by fentanyl. This study was conducted at only one concentration of midazolam and fentanyl, and thus this may not truly reflect the entire interaction profile between these two drugs for amnesia. From the intrinsic properties of the benzodiazepines, amnesia will occur at much lower hypnotic concentrations of the drug. It is not established if even lower concentrations are needed when they are used in combination with other hypnotics or opioids. It is even less well established if combinations of non-benzodiazepine hypnotics and opioids provide an enhanced amnesic effect. With several reports of awareness appearing recently in the literature during TIVA, it would be best to assume that absence of awareness and recall can only be assured if an adequate concentration of the hypnotic rather than the opioid (which have limited amnesic properties) is maintained. Until these data are available it is recommended that the concentration of the volatile anaesthetic or propofol should not be titrated much below its MACawake value because of the risk of awareness and recall. With the advent of bispectral monitoring, this monitor may act as a guide to the dosing of the hypnotic.

From the above it is clear that drug interactions vary considerably for the same combination depending on what effect is being measured. There is little advantage for a synergistic interaction for loss of consciousness if the same drug combination results in an even greater synergistic cardiovascular depression. It is well known that opioids plus a benzodiazepine produce significant decreases in blood pressure [37] and also tend to provide more profound respiratory depression than when used alone. Thus, although various combinations of the intravenous anaesthetics will allow less drug to be given for the same anaesthetic endpoint, it is important to evaluate the effects of the drug combination on other physiological variables so that the optimal dosing combination to both provide adequate anaesthesia and homeostasis is used.

Intravenous anaesthetics have numerous pharmacological properties other than just loss of consciousness or analgesia. Thus, not only do intravenous anaesthetics alter the effects of other intravenous anaesthetics, but they may also alter the effects of other drugs used during an anaesthetic. In general, unlike the volatile anaesthetics, intravenous hypnotics and opiates do not potentiate neuromuscular blockers. Intravenous anaesthetics have profound effects on the cardiovascular control mechanisms as well as direct effects on the heart and blood vessels. It is, therefore, not unexpected for the effects of vasoactive drugs to be altered in the presence of intravenous anaesthetics. Similarly, intravenous anaesthetics alter cerebral metabolism, blood flow and excitability, and thus, similarly, they are sometimes used to potentiate the effects of other centrally acting drugs.

It is also important to be aware of the drugs the patient is receiving prior to anaesthesia as these drugs may result in a pronounced pharmacodynamic interaction with the intravenous anaesthetics. Such interactions are poorly documented but it is well known that centrally acting drugs (e.g. clonidine or antihistamines) alter MAC. These drugs are just as likely to alter the requirements of the intravenous anaesthetics. In addition, chronic use of opiates, benzodiazepines, alcohol, etc., result in tolerance that similarly will alter the pharmacodynamic effect of an intravenous anaesthetic in these patients. It also appears that combining drugs may alter the onset of tolerance to opiates and benzodiazepines. Acute analgesic tolerance could be induced in volunteers when a steady concentration of fentanyl was administered. Similarly, acute tolerance could also be induced with midazolam for psychomotor function. However, when the drugs were used in combination, tolerance to neither the analgesia nor psychomotor function was observed [38].

Anaesthesia involves the administration of numerous drugs which are required for the anaesthetic state. Patients presenting for anaesthesia may also be taking a variety of drugs. The net result is that there is a marked potential for drug interactions during anaesthesia which may occur as a result of a pharmacodynamic, a pharmacokinetic, or a pharmaceutical interaction [1,2]. The pharmacodynamic interactions between anaesthetic drugs appear to be clinically most important in providing the anaesthetic state and with which the anaesthetist needs to become familiar. It is, therefore, important to incorporate the knowledge of drug interactions to provide optimal dosing regimens for patients presenting for anaesthesia.


1 Eger EI, Saidman LJ, Brandstater B. Minimum alveolar anesthetic concentration: a standard of anesthetic potency. Anesthesiology 1965; 26: 756-763.
2 Sear JW. Drug interactions and adverse reactions. In: Baillière's Clinical Anaesthesiology, London: Baillière Tindall, 1991; 5: 703-733.
3 Mueller RA, Lundberg DBA. Philosophy and theory. In: Drug Interactions for Anesthesiology (2nd ed), New York: Churchill Livingstone, 1992: 3-19.
4 Kissin I, Mason JO III, Bradley EL. Morphine and fentanyl interactions with thiopental in relation to movement response to noxious stimulation. Anesth Analg 1986; 65: 1149-1154.
5 Roytblat L, Katz J, Rozentsveig V, Gesztes T. Bradley EL Jr., Kissin I. Anaesthetic interaction between thiopentone and ketamine. Eur J Anaesthesiol 1992; 9: 307-312.
6 Daghero AM, Bradley EL Jr, Kissin I. Midazolam antagonizes analgesic effect of morphine in rats. Anesth Analg 1987; 66: 944-947.
7 Kissin I, Mason JO III, Bradley EL Jr. Morphine and fentanyl hypnotic interactions with thiopental. Anesthesiology 1987; 67: 331-335.
8 Kissin I, Mason JO III, Bradley EL Jr. Pentobarbital and thiopental anesthetic interactions with midazolam. Anesthesiology 1987; 67: 26-31.
9 Kissin I, Brown PT, Bradley EL, Robinson CA, Cassady JL. Diazepam-morphine hypnotic synergism in rats. Anesthesiology 1989; 70: 689-694.
10 Kissin I, Vinik HR, Castillo R, Bradley EL. Alfentanil potentiates midazolam-induced unconsciousness in subanalgesic doses. Anesth Analg 1990; 71: 65-69.
11 Kissin I, Brown PT, Bradley EL. Sedative and hypnotic midazolam-morphine interactions in rats. Anesth Analg 1990; 71: 137-143.
12 Kissin I, Brown PT, Bradley EL. Morphine and fentanyl anesthetic interactions with diazepam: relative antagonism in rats. Anesth Analg 1990; 71: 236-241.
13 Kissin I, Vinik HR, Bradley EL Jr. Midazolam potentiates thiopental anesthetic induction in patients. J Clin Anesth 1991; 3: 367-370.
14 Mehta D, Bradley EL Jr., Kissin I. Effect of alfentanil on hypnotic and antinociceptive components of thiopental sodium anesthesia. J Clin Anesth 1991; 3: 280-284.
15 Tverskoy M, Fleyshman G, Ezry J, Bradley EL, Kissin I. Midazolam-morphine sedative interaction in patients. Anesth Analg 1989; 68: 282-285.
16 Tverskoy M, Ben-Shlomo I, Ezry J, Finger J, Fleyshman G. Midazolam acts synergistically with methohexitone for induction of anaesthesia. Br J Anaesth 1989; 63: 109-112.
17 Tverskoy M, Fleyshman G, Bradley EL, Kissin I. Midazolamthiopental anesthetic interaction in patients. Anesth Analg 1988; 67: 342-345.
18 Vinik HR, Bradley EL, Kissin I. Midazolam-alfentanil synergism for anesthetic induction in patients. Anesth Analg 1989; 69: 213-218.
19 Short TG, Plummer JL, Chui PT. Hypnotic and anaesthetic interactions between midazolam, propofol and alfentanil. Br J Anaesth 1992; 69: 162-167.
20 McClune S, McKay AC, Wright PMC, Patterson CC, Clarke SJ. Synergistic interaction between midazolam and propofol. Br J Anaesth 1992; 69: 240-245.
21 Short TG, Galletly DC, Plummer JL. Hypnotic and anaesthetic action of thiopentone and midazolam alone and in combination. Br J Anaesth 1991; 66: 13-19.
22 Short TG, Chui PT. Propofol and midazolam act synergistically in combination. Br J Anaesth 1991; 67: 539-545.
23 Naguib M, Sari-Kouzel A. Thiopentone-propofol hypnotic synergism in patients. Br J Anaesth 1991; 67: 4-6.
24 Vinik HR, Bradley EL Jr, Kissin I. Triple anesthetic combination: propofol, midazolam, alfentanil. Anesth Analg 1994; 78: 354-358.
25 Glass PSA, Shafer SL, Jacobs JR, Reves JG. Intravenous drug delivery systems. In: Miller R, ed. Anesthesia 4th ed. New York: Churchill Livingstone, 1994; 389-416.
26 McEwan A, Smith C, Dyar O, Goodman D, Glass PS. Isoflurane MAC reduction by fentanyl. Anesthesiology 1993; 78: 864-869.
27 Westmoreland C, Sebel PS, Groper A, Hug CC Jr. Reduction of isoflurane MAC by fentanyl or alfentanil. Anesthesiology 1992; 77: A394.
28 Brunner MD, Braithwaite P, Jhaveri R et al. The MAC reduction of isoflurane by sufentanil. Br J Anaesth 1994; 72: 42-46.
29 Lang E, Kapila A, Shlugman D, Hoke JF, Sebel PS, Glass PSA. Reduction of isoflurane Minimal Alveolar Concentration by remifentanil. Anesthesiology 1996; 85: 721-728.
30 Glass PSA, Hardman D, Kamiyama Y et al. Preliminary pharmacokinetics and pharmacodynamics of an ultra-short-acting opioid: remifentanil (GI87084B). Anesth Analg 1993; 77: 1031-140.
31 Westmoreland CL, Hoke JF, Sebel PS, Hug CC, Muir KT. Pharmacokinetics of remifentanil (GI87084B) and its metabolite (GI90291) in patients undergoing elective inpatient surgery. Anesthesiology 1993; 79: 893-903.
32 Egan TD, Lemmens HJM, Fiset P et al. The pharmacokinetics of the new short-acting opioid remifentanil (GI87084B) in healthy adult male volunteers. Anesthesiology 1993; 80: 881-892.
33 Smith C, McEwan AI, Jhaveri R et al. The interaction of fentanyl on the Cp50 of propofol for loss of consciousness and skin incision. Anesthesiology 1994; 81: 820-828.
34 Telford RJ, Glass PSA, Goodman D, Jacobs JR. Fentanyl does not alter the "sleep" plasma concentration of thiopental. Anesth Analg 1992; 75:523-529.
35 Vuyk J, Lim T, Engbers FHM, Burm AGL, Vletter AA, Bovill JG. The pharmacodynamic interaction of propofol and alfentanil during lower abdominal surgery in female patients. Anesthesiology 1995; 83: 8-22.
36 Hughes MC, Glass PSA, Jacobs JR. Context-sensitive half-time in multicompartment pharmacokinetic models for intravenous anesthetic drugs. Anesthesiology 1992; 76: 334-341.
37 Reves JG, Kissin I, Fournier SE, Smith LR. Additive negative inotropic effect of a combination of diazepam and fentanyl. Anesth Analg 1984; 63: 93-100.
38 Glass P, Camporesi EM, Farinelli LA. Acute tolerance to fentanyl and midazolam. Anesth Analg 1992; 74: S111.

Section Description

Seventh International Symposium on Intravenous Anaesthesia, Lausanne, Switzerland, 2-3 May 1997

This publication is supported by grants from various pharmaceutical companies. The views in this publication are those of the authors and not necessarily those of supporting companies. Drugs and administration techniques referred to should only be used as recommended in the manufacturers' prescribing information.


Pharmacodynamics; Interactions (drug); Antagonists; Anaesthetics Intravenous

© 1997 European Society of Anaesthesiology