THE anesthesia research community has contributed enormously to the development of modern pharmacokinetic and pharmacodynamic methodologies. For example, pharmacokinetic studies defined the plasma concentration profile during the first minute after bolus administration (demonstrating the flaw in assuming that plasma concentration peaked at time 0, then decreased monotonically2
); other studies disproved relevance of the traditional “elimination half-life” (demonstrating that duration of administration of anesthetic drugs markedly influenced washout after administration3
). Pharmacodynamic studies introduced the concept of an effect compartment4
, that the effect of a drug might not be explained by its plasma concentration), thereby reconciling the observation that shortly after bolus administration, plasma concentrations are decreasing yet the effect of most drugs is increasing. Yet, as science evolves, new methodologies might be flawed, yielding specious conclusions. Such could have been the case in an article in the current issue of Anesthesiology,5
had the authors not reflected on the implications of their study.
Several notable examples of new methodologies leading to flawed conclusions come to mind. In 1980, Toner et al.6
administered single bolus doses of thiopental to healthy subjects and measured thiopental concentrations. Larger doses were associated with higher venous concentrations at awakening, leading the authors to conclude that there was acute tolerance to thiopental. Subsequently, Hudson et al.7
demonstrated that this conclusion was an artifact—modeled brain concentrations of thiopental did not vary with dose, i.e.
, there was no evidence to support acute tolerance. Thus, the use of venous concentrations as a biomarker for effect compartment concentration led to a flawed conclusion.
A second example involved a similar conclusion: that different doses of a muscle relaxant resulted in different concentrations producing 50% twitch depression (C50
). In that case, Bergeron et al.8
administered cisatracurium and applied accepted pharmacokinetic–pharmacodynamic models to estimate C50
. They reported a 1.5-fold change in C50
over a 4-fold range of doses. I thought that their result could be an artifact of the analysis method; specifically, their flawed assumption that Cp peaked instantly, and then decreased monotonically, would have a different impact depending on the magnitude of the dose administered (because larger doses suppressed twitch markedly more rapidly than smaller doses). Paul and I9
used simulation to demonstrate that the results of Bergeron et al.
were likely to be artifactual. Ironically, the importance of proper characterization of the plasma concentration profile during the first few minutes after bolus administration had been championed earlier by some of Bergeron's coauthors.10
Both of these examples (and there are more) show that applying accepted methodologies may yield spurious outcomes, especially when the new studies “take it to the limit.” Struys et al. encounter a similar situation. They were interested in the inconsistent published values for the half-life for effect site equilibration (t½ke0) for propofol. Previous studies suggested that this half-life differed when propofol was administered by bolus compared with rapid infusion. Was this an artifact? Or did it represent a true phenomenon that should be accounted for if one administered propofol by target-controlled infusion?
To address this question, Struys et al.
administered 2.5 mg/kg propofol as either a bolus or a rapid infusion (duration of 1, 2, or 3 min) to healthy subjects who received no other drugs; spectral edge encephalograph was measured as the effect. Instead of measuring plasma concentrations of propofol, they estimated plasma concentrations using a “validated” set of pharmacokinetic parameters. Their initial analysis indicated that equilibration half-life differed twofold between the two modes of administration. The authors could have accepted this as a plausible result: Ludbrook et al.11
had reported that high concentrations of propofol constrict cerebral vessels, which might delay blood–brain equilibration.
Fortunately, Struys et al.
did not stop there. When challenged that their results could be artifactual, a result of misspecifying the plasma concentration profile during the first minutes after bolus administration, Struys et al.
attempted to validate their plasma concentration estimates by sampling propofol frequently after bolus administration in a new cohort. To their surprise, measured plasma concentrations during the first minutes were severalfold lower than the values they estimated. Perhaps they should not have been surprised: Data from Doufas et al.12
suggest a similar phenomenon.
Struys et al. then evaluated additional models to determine the impact of these new plasma concentration values. Interestingly, most of the difference in equilibration half-life between the bolus and infusion groups disappeared. Thus, their initial conclusion—that mode of administration affected equilibration half-life—was no longer tenable. Had Struys et al. not studied the additional cohort, they would have published an article containing a flawed conclusion, creating the opportunity for a future study to refute their findings. Alternatively, a flawed conclusion might have gone unchallenged, potentially for a lengthy period.
Science evolves constantly. We benefit from investigators taking studies to the limits of the available methodologies. And we also benefit when mistakes are made, acknowledged, and corrected.
Dennis M. Fisher, M.D.
“P Less Than” Company, San Francisco, California. firstname.lastname@example.org
1. Henley D, Frey G, Meisner R: Take it to the limit. Asylum Records 1975
2. Henthorn TK, Avram MJ, Krejcie TC: Intravascular mixing and drug distribution: The concurrent disposition of thiopental and indocyanine green. Clin Pharmacol Ther 1989; 45:56–65
3. Hughes MA, Glass PS, Jacobs JR: Context-sensitive half-time in multicompartment pharmacokinetic models for intravenous anesthetic drugs. Anesthesiology 1992; 76:334–41
4. Sheiner LB, Stanski DR, Vozeh S, Miller RD, Ham J: Simultaneous modeling of pharmacokinetics and pharmacodynamics: Application to d-tubocurarine. Clin Pharmacol Ther 1979; 25:358–71
5. Struys MMRF, Coppens MJ, De Neve N, Mortier EP, Doufas AG, Van Bocxlaer HFP, Shafer SL: Influence of administration rate on propofol plasma–effect site equilibration. Anesthesiology 2007; 107:386–96
6. Toner W, Howard PJ, McGowan WA, Dundee JW: Another look at acute tolerance to thiopentone. Br J Anaesth 1980; 52:1005–8
7. Hudson RJ, Stanski DR, Saidman LJ, Meathe E: A model for studying depth of anesthesia and acute tolerance to thiopental. Anesthesiology 1983; 59:301–8
8. Bergeron L, Bevan DR, Berrill A, Kahwaji R, Varin F: Concentration–effect relationship of cisatracurium at three different dose levels in the anesthetized patient. Anesthesiology 2001; 95:314–23
9. Paul M, Fisher DM: Pharmacodynamic modeling of muscle relaxants: Effect of design issues on results. Anesthesiology 2002; 96:711–7
10. Donati F, Varin F, Ducharme J, Gill SS, Theoret Y, Bevan DR: Pharmacokinetics and pharmacodynamics of atracurium obtained with arterial and venous blood samples. Clin Pharmacol Ther 1991; 49:515–22
11. Ludbrook GL, Visco E, Lam AM: Propofol: Relation between brain concentrations, electroencephalogram, middle cerebral artery blood flow velocity, and cerebral oxygen extraction during induction of anesthesia. Anesthesiology 2002; 97:1363–70
12. Doufas AG, Bakhshandeh M, Bjorksten AR, Shafer SL, Sessler DI: Induction speed is not a determinant of propofol pharmacodynamics. Anesthesiology 2004; 101:1112–21
© 2007 American Society of Anesthesiologists, Inc.