To the Editor:—
We read with great interest the recent report by Gyulai et al. 1regarding the ability of isoflurane to enhance the in vivo binding of 11C-flumazenil to the GABAA–benzodiazepine receptors in the human brain. This work demonstrates the feasibility of using brain imaging technology to help bridge the gap in knowledge that exists between the in vitro and the in vivo worlds of anesthesia research. This effort complements our finding that propofol's in vivo regional cerebral metabolic depressant effects correlate well with the reported regional ex vivo receptor density data of human benzodiazepine binding sites. 2In contrast, however, we also found that the regional cerebral metabolic depressant effects of isoflurane did not appear to correlate with the benzodiazepine receptor density data. 2The present work by Gyulai et al. suggests that the regional cerebral metabolic depression caused by isoflurane may indeed involve the GABA receptor, but perhaps in a less regionally specific manner than that of propofol.
However, before accepting this idea, a technical issue relevant to the emerging field of brain imaging applied to anesthesia research needs consideration. An issue that may play a role in the results presented by Gyulai et al. is the anesthetic-induced alteration of the arterial plasma drug concentration-versus -time relationship in the minutes after rapid intravenous administration of the tracer (the front-end kinetics). 3The experimental group in the study of Gyulai et al. was anesthetized with isoflurane, and the subjects’ arterial blood pressures were maintained at control values by infusing phenylephrine. Isoflurane and phenylephrine, by the combined effect of depressing cardiac output, may have increased the fraction of the dose of tracer to which the brain was exposed. 4
One measure of tissue drug exposure is the area under the arterial plasma concentration-versus -time curve (AUC) for the time during which the concentration of the drug remains above some threshold value. We have previously demonstrated that in dogs anesthetized with isoflurane (1.7 MAC), the AUC0–3minfor antipyrine, a marker of lipophilic drug disposition, more than doubled compared to the same animals while awake. 5Likewise, when awake dogs were treated with an infusion of phenylephrine in a dose sufficient to double the baseline calculated systemic vascular resistance, the AUC0–3minincreased by 75% relative to placebo-treated awake animals. 6We discovered that the increased arterial concentrations of this lipophilic marker were due to a relative increase in the proportion of cardiac output not involved in the distribution of drug to peripheral tissues. The nondistributive blood flow acts as a “pharmacokinetic shunt.” This pharmacokinetic shunt could have significantly increased the amount of tracer delivered to the brain in the isoflurane–phenylephrine condition in the study of Gyulai et al. , making it appear as if the isoflurane had done something in the brain to increase tracer binding.
Gyulai et al. did, however, report regional differences in apparent binding and that “no significant differences were observed in nonreceptor ligand binding between the two experimental conditions in either experimental group. This would indicate that the observed changes in binding were not confounded by altered ligand delivery.”1Nevertheless, perhaps the authors could have included a control condition in which a few subjects would have received only a phenylephrine infusion. The increased nondistributive blood flow caused by phenylephrine results in an increase in AUC similar to that seen with isoflurane. 6Thus, if such an active control experiment were to reveal increased tracer binding relative to the saline-treated controls, then perhaps one might conclude that some of the published isoflurane effect should be attributed to an anesthetic-induced change in tracer kinetics, rather than to a central GABAergic effect of isoflurane itself.
In any event, the work by Gyulai et al. provides further evidence that functional brain imaging technology has a role to play in helping us elucidate mechanisms of anesthetic action, and we suggest that some of the work left to be done might involve obtaining a better understanding of how our anesthetic agents interact with the techniques being used to study them.