Vol. 284, Issue 2, 460-466, February 1998

Pharmacokinetic-Pharmacodynamic Modeling of the Anticonvulsant and Electroencephalogram Effects of Phenytoin in Rats

O. E. Della Paschoa, J. W. Mandema, R. A. Voskuyl and M. Danhof

Leiden/Amsterdam Center for Drug Research, Division of Pharmacology, University of Leiden, University of Leiden, P.O. Box. 9503, 2300 RA Leiden, The Netherlands (O.E.D.P., M.D.), Stanford University School of Medicine, Department of Anesthesia, Palo Alto, California (J.W.M.) and Instituut voor Epilepsiebestrijding, Heemstede, The Netherlands (R.A.V.)

In this study a pharmacokinetic-pharmacodynamic model is proposed for drugs with nonlinear elimination kinetics. We applied such an integrated approach to characterize the pharmacokinetic-pharmacodynamic relationship of phenytoin. In parallel, the anticonvulsant effect and the electroencephalogram (EEG) effect were used to determine the pharmacodynamics. Male Wistar-derived rats received a single intravenous dose of 40 mg · kg¹ phenytoin. The increase in the threshold for generalized seizure activity (TGS) was used as the anticonvulsant effect and the increase in the total number of waves in the 11.5 to 30 Hz frequency band was taken as the EEG effect measure. Phenytoin pharmacokinetics was described by a saturation kinetics model with Michaelis-Menten elimination. V_max and K_m values were, respectively, 386 ± 31 µg · min¹ and 15.4 ± 2.2 µg · ml¹ for the anticonvulsant effect in the cortical stimulation model and 272 ± 31 µg · min¹ and 5.9 ± 0.7 µg · ml¹ for the EEG effect. In both groups, a delay to the onset of the effect was observed relative to plasma concentrations. The relationship between phenytoin plasma concentrations and effect site was estimated by an equilibration kinetics routine, yielding mean k_e0 values of 0.108 and 0.077 min¹ for the anticonvulsant and EEG effects, respectively. The EEG changes in the total number of waves could be fitted by the sigmoid E_max model, but E_max values could not be estimated for the nonlinear relationship between concentration and the increase in TGS. An exponential equation (E = E₀ + Bⁿ · Cⁿ) derived from the sigmoid E_max model was applied to describe the concentration-anticonvulsant effect relationship, under the assumption that E_max values cannot be reached within acceptable electric stimulation levels. This approach yielded a coefficient (B) of 2.0 ± 0.4 µA · ml · µg¹ and an exponent (n) of 2.7 ± 0.9. The derived EC₅₀ value of 12.5 ± 1.3 µg · ml¹ for the EEG effect coincides with the "therapeutic range" in humans.