Pharmacokinetic-pharmacodynamic modeling the hypnotic effect of sevoflurane using the spectral entropy of the electroencephalogram.

Spectral entropy is a new electroencephalogram (EEG)-derived parameter that may be used to model the pharmacokinetic-pharmacodynamic (PKPD) effects of general anesthetics. In the present study we sought to derive a PKPD model of the relationship between sevoflurane concentration and spectral entropy of the EEG. We collected spectral entropy data during increasing and decreasing sevoflurane anesthesia from 20 patients. The first cycle consisted of induction and lightening phases with no supplemental medications. An effect-site compartment and inhibitory E(max) model described the relation between sevoflurane concentration and spectral entropy. PKPD parameters were derived from the full cycle and separately from the increasing and decreasing stages. The second anesthetic cycle consisted of a redeepening phase only and included airway manipulation and routinely administered adjunctives. PKPD data obtained from the first cycle were used to predict second cycle entropy changes. There was a consistent relationship between effect-site sevoflurane concentration and spectral entropy (median absolute weighted residual = 11.6%). For complete first-cycle response entropy (mean +/- sd): T1/2 K(eo) = 2.4 +/- 1.5 min, gamma = 5.9 +/- 2.3, EC50 = 1.7 +/- 0.3. We found significant differences between gamma values when the sevoflurane concentration was increasing (61.1 +/- 55.2) compared with the decreasing part of the cycle (5.7 +/- 2.8). Above an effect-site concentration of 3%, spectral entropy of the EEG is unresponsive to further increases in sevoflurane concentration. The effect-compartment inhibitory E(max) model accurately describes the relation between sevoflurane concentration and spectral entropy of the EEG. Spectral entropy decreases with increasing sevoflurane concentrations up to 3%. The steepness of the dose-response curve varies between phases of increasing and decreasing anesthetic concentrations.