00/0
THE JOURNAL OF PHARMACOLOGY AND EXPERIMENTAL THERAPEUTICS Vol. 284, No. 2
Copyright 1998 by The American Society for Pharmacology and Experimental Therapeutics Printed in U.S.A.
JPET 284:460 466, 1998
Phenytoin (diphenylhydantoin) has been used extensively anticonvulsant properties of antiepileptic drugs in preclinical
in the treatment of seizure disorders. Although its pharma- investigation, such as the kindling and maximal electroshock
cokinetics has been well reported both in animals and in models, have limitations which prohibit the assessment of
humans (Olanow and Finn, 1981; Bauer and Blouin, 1983; the PK-PD relationship (Rundfeldt et al., 1990; Rundfeldt
Shavit et al., 1984; Jones and Wimbish, 1985; Levine and and Loscher, 1993; Mulzac and Scott, 1993; Dimmock and
Chang, 1990) the concentration-anticonvulsant effect profile Baker, 1994). A major limitation is the impossibility to de-
of phenytoin is not thoroughly understood. Consequently, termine repeatedly the anticonvulsant effect intensity within
prediction of the accurate PK-PD relationship for both clini- individual rats.
cal and preclinical scopes is still a difficult undertaking. This Furthermore, approaches in which the use of antagonists
can be attributed largely to the lack of adequate quantitative might fully explain the effect of the agonist are not applicable
measures of the effect of AED in vivo (Danhof et al., 1992b). in vivo because of the mechanism of action of phenytoin. The
The nonlinear pharmacokinetics of phenytoin represents an neuropharmacological and biochemical effects of phenytoin
additional complicating factor (Theodore, 1992). are vast. Some of the reported actions of phenytoin in various
To date, the numerous models developed to delineate the systems include: (1) changes in the conductance of ionic chan-
nels (Jones and Wimbish, 1985); (2) inhibition of calcium
Received for publication March 10, 1997. uptake and calcium-dependent protein phosphorylation
ABBREVIATIONS: CSM, cortical stimulation model; E0, base-line effect; Emax, maximal effect; EC50, concentration at half maximal effect; EEG,
electroencephalogram; fu, free fraction; Km, Michaelis constant; PK-PD, pharmacokinetic-pharmacodynamic; TGS, threshold for generalized
seizure activity; TLS, threshold for localized seizure activity; Vd, volume of distribution; Vmax, maximum metabolic rate; TNW, total number of
waves.
460
1998 PK-PD Modeling of Phenytoin 461
(Twombly et al., 1988); (3) elevation of membrane potential samples were then centrifuged for 10 min at 5000 rpm and plasma
and changes in the amplitudes of synaptic potentials; (4) (50 or 100 ml) separated and stored at 230C until analysis.
suppression of bursting activity; (5) increase in cortical levels
of g-aminobutyric acid (Griffith and Taylor, 1988). EEG Measurements and Cortical Stimulation
With respect to the pharmacokinetics, it is important to Group I. The output from bipolar leads was continuously moni-
consider that the decay in plasma concentration is not linear. tored by a Nihon Kohden EEG recorder. During the course of the
Because the rate at which phenytoin is hydroxylated is dose experiment, animals were kept in motion in a slow-speed rotating
dependent, elimination can follow both first and zero-order drum (10 rph) to control the vigilance level. Base-line activity was
processes, depending on the concentration range. A model monitored for 30 min. After drug administration, EEG recordings
with Michaelis-Menten elimination is therefore required to were continued for 5 hr. The frontocentral lead on the left hemi-
describe the pharmacokinetics of phenytoin (Gibaldi and Per- sphere was subjected to on-line aperiodic analysis for quantification
rier, 1982). of the effect. The aperiodic analysis algorithm calculates the ampli-
tudes of each EEG signal on a wave-by-wave basis. The analyzed
In addition, one has to consider that the bulk of PK-PD
EEG data were stored on a magnetic floppy disk. The drug-induced
modeling theory and methodology pertains to linear pharma- change in the total number of waves in the 12.0- to 30.0-Hz frequency
cokinetics with specific drug-receptor interaction systems band was applied as effect measure.
(Danhof et al., 1992a, 1993). Mathematical models and meth- Group II. The seizure thresholds were determined as described
ods for such nonlinear systems are much less developed previously (Voskuyl et al., 1989). Convulsive activity was induced by
(Ritschel and Hussain, 1984; van Rossum and Burgers, 1984; a single train of bipolar pulses (total pulse duration, 2 msec, 50 Hz)
Phenytoin Assay
Methods
Samples were analyzed by the high-performance liquid chroma-
Study Design tography technique essentially in the manner outlined by Ratnaraj et
al.(1989). A solution of 1.5 mg mephenytoin (internal standard) in
The study protocol was approved by the Ethical Committee for
150 ml acetonitrile was added to a 50- or 100-ml plasma sample. To
Animal Experimentation. The anticonvulsant activity and the EEG
separate the organic phase, 100 ml saturated NaH2PO4 solution were
effect of phenytoin were determined in two groups of rats according
added and then whirl-mixed for 15 sec. After 10 min centrifugation
to a parallel group design. Two groups of male Wistar-derived rats
at 5000 rpm, 75 ml were transferred from the supernatant and 25 ml
(Sylvius Laboratory Breeding Facility, Leiden, The Netherlands)
were injected into the chromatographic system. The mobile phase
(225300 g) were used throughout the study. The animals were
consisted of a mixture of 0.067 M phosphate buffer (pH 5 5.6) and
housed individually in plastic cages under constant temperature
acetonitrile in a 70:30 ratio with a flow rate of 1.7 ml z min21. The
(21C) and 12-hr light/dark cycle. Laboratory chow (Standard Labo-
high-performance liquid chromatography system consisted of a Kra-
ratory Rat, Mouse and Hamster Diets, RMH-TM, Hope Farms,
tos solvent delivery system, a WISP-710B automatic sample injector
Woerden, The Netherlands) and water were available ad libitum,
and a Spectroflow 757 Kratos spectrophotometer (wave length, 215
except during the experimental procedures.
nm). The analytical column was a 25 cm 3 4.6 mm i.d. Altex column
filled with Ultrasphereo-ODS 5 m. Data processing was performed by
Surgical Procedure a Chromatopack C-R3A reporting integrator. Phenytoin retention
Seven chronic cortical EEG electrodes were implanted in the skull time was 8.0 min. Within-day precision was 1.7% for a 10 mg z ml21
of the animals (Mandema and Danhof, 1990) 1 week before the control sample (n 5 8). Limit of detection was 0.25 mg z ml21 and the
experiments for the measurement of EEG signals (group I). Implan- assay was linear in the range from 1 to 100 mg z ml21.
tation of two permanent electrodes over the motor area of the fron-
toparietal cortex (Voskuyl et al., 1989) allowed the assessment of the Protein Binding
anticonvulsant effect (group II). One day before the measurements,
indwelling cannulae were implanted into the right jugular vein (for Residual blood was collected by aorta puncture after completion of
drug administration) and femoral artery (for blood sampling). the experiments and centrifuged for 10 min at 5000 rpm. Plasma was
separated and stored at 230C until assay. The protein binding was
determined for each individual animal by ultrafiltration at 37C,
Drug Dosage, Blood Sampling and Pharmacokinetics with use of the Amicon Micropartition System (Amicon Division,
A 40 mg z kg21 dose of phenytoin was infused intravenously at a Danvers, MA). Three 0.5-ml aliquots of a plasma sample were spiked
rate of 0.1 ml z min21 for 5 min. Phenytoin sodium (Sigma Chemical with phenytoin to a concentration of 10, 50 and 100 mg z ml21. Total
Co., St. Louis, MO) was dissolved in water alkalinized with 0.1 N plasma concentration was measured by the described method in a
sodium hydroxide. Arterial blood samples (100 or 200 ml) were col- 50-ml sample of the spiked plasma. Separation of free drug from
lected in heparinized tubes before and 5, 10, 15, 20, 30, 40, 60, 90, protein-bound drug was carried out at 37C by filtration of a 400-ml
120, 150, 180, 240 and 300 min after drug administration. Blood plasma through a YMT ultrafiltration membrane (Amicon) at 1090 3
462 Paschoa et al. Vol. 284
g for 10 min. The ultrafiltrate was then analyzed for free drug where E(Ce) is the observed effect at concentration C, E0 is the
concentrations. base-line effect value, Emax is the maximal effect, EC50 is the con-
centration at half-maximal effect and n is a constant expressing the
Data Analysis shape of the concentration-effect relationship.
Finally, based on the sigmoid Emax model we have derived an
The pharmacokinetics and pharmacodynamics of phenytoin were equation to describe the exponential profile of the anticonvulsant
quantified for each individual rat. First, a two-compartment model effect. This was performed under the assumption that both EC50 and
with Michaelis-Menten elimination was used to describe the plasma Emax values are very high and cannot be determined experimentally,
concentration-time profile (Gibaldi and Perrier, 1982): without lesion of the brain or animal harm. When drug concentra-
dC 1 V max z C 1 R tions are small in relation to EC50, the equation tends to a constant
52 1 2 k 12 z C 1 1 k 21 z C 2 (1) value in the denominator. In practice, the combined parameter Emax/
dt Km 1 C1 V1 EC50n may be used to describe drug effect in the range observed:
dC 2
5 k 12 z C 1 2 k 21 z C 2 (2) E max n
dt E ~ C e! 5 E 0 1 n zC when C ,, EC50 (6)
EC50
with the error model
Assuming that Emax/EC50n 5 Bn, the equation can be rewritten as:
log~ C m! 5 log~ C pi! 1 i (3)
n n
where dC1/dt is the rate of decline of drug concentration at time t, V1 E ~ C e! 5 E 0 1 B z C (7)
Pharmacokinetics
Vmax (mg z min21) 272 6 31 386 6 31b
Km (mg z ml21) 5.9 6 0.7 15.4 6 2.2c
V1 (ml) 406 6 32 184 6 13d
k12 (min21) 0.197 6 0.018 0.255 6 0.019c
k21 (min21) 0.060 6 0.004 0.072 6 0.002b
fu (%) 30.2 6 2.4 25.3 6 3.0
Pharmacodynamics
ke0 (min21) 0.077 6 0.020 0.108 6 0.017
E0 (TNW or mA) 6.0 6 0.2 685 6 14
Emax (TNW) 2.6 6 0.3
B 2.0 6 0.4
(mA z ml z mg21)
Hill factor 4.6 6 0.9 n 5 2.7 6 0.9
EC50 (mg z ml21) 12.6 6 1.3
a
n $ 7; mean 6 S.E. Statistical significance: b
P , .05; c
P , .01; d
P , .001
(ANOVA).
Results
Pharmacokinetics
The kinetic disposition of phenytoin followed a two-com-
partment model with Michaelis-Menten elimination. The av-
eraged concentration profile is plotted against time in figure
2. Vmax and Km presented somewhat higher values in group
II (table 1). Because protein binding was similar in both
groups, the lower values of volume of distribution apparently
compensated for such differences, without additional conse-
quences for biophase equilibration and pharmacodynamics. Fig. 3. The hysteresis between plasma concentration and EEG effect of
an individual rat is normalized after minimization with a monoexponen-
tial conductance function. The same is observed for the hysteresis be-
Hysteresis tween plasma concentration and the anticonvulsant effect (TGS). The
The effect onset was characterized by a temporal delay arrow indicates the time sequence of the observations.
relative to plasma concentrations in both groups. As shown
in figure 3, the hysteresis was minimized successfully by the cannot be reached at the administered dose. The exponential
link model. ke0 values of the same magnitude were obtained equation 7 was used to describe the nonlinear profile of the
for the EEG and anticonvulsant effects. anticonvulsant effect (fig. 5). The pharmacokinetic and phar-
macodynamic parameters generated from the PK-PD model-
Pharmacodynamics ing are summarized in table 1.
Group I. Intravenous administration of phenytoin in-
duced changes in both amplitudes and TNW in the beta
frequency band (12.0 30.0 Hz) of the EEG spectral power.
Discussion
The increase in TNW was selected as the EEG effect measure In this study we demonstrate the application of an inte-
(fig. 4). Because an Emax value was reached, the concentra- grated PK-PD approach to characterize the biophase equili-
tion-EEG effect relationship could be characterized by the bration kinetics and the concentration-effect relationship of
sigmoid Emax model (fig. 5). phenytoin both in the EEG and in the cortical stimulation
Group II. The anticonvulsant effect of phenytoin was re- models.
flected by the elevation of the TGS without any significant The nonlinear pharmacokinetic profile of phenytoin, de-
alteration in the TLS. As depicted in figure 4, the effect-time scribed by a two-compartment model with Michaelis-Menten
course profile followed a pattern remarkably similar to the elimination, presented high variability. Our data agree with
changes observed in the EEG. However, fitting these data to published data, however (Jones and Wimbish, 1985; Itoh et
the sigmoid Emax model was not possible. It seems that Emax al., 1988). Km (Michaelis constant) and Vmax (maximum met-
464 Paschoa et al. Vol. 284
understanding about the nature of the observed effects and Mandema JW, Kuck MT and Danhof M (1992) Differences in intrinsic efficacy of
benzodiazepines are reflected in their concentration-EEG effect relationship. Br J
the underlying concentration-effect relationship. Application Pharmacol 105:164 170.
of this methodology represents an important step forward in Mandema JW and Danhof M (1990) Pharmacokinetic-pharmacodynamic modeling of
the comprehension of the pharmacodynamics of drugs with the central nervous system effects of heptabarbital using aperiodic EEG analysis.
J Pharmacokinet Biopharm 18:459 481.
nonlinear pharmacokinetics. Mandema JW and Danhof M (1992) Electroencephalogram effect measures and
relationships between pharmacokinetics and pharmacodynamics of centrally act-
ing drugs. Clin Pharmacokinet 23:91215.
Acknowledgments
Mulzac D and Scott KR (1993) Profile of anticonvulsant activity and minimal toxicity
The authors thank K. B. Postel-Westra and M. Langemeijer for of methyl 4-[(p-chlorophenyl)amino]-6-methyl-2-oxo-cyclohex-3-en-1-oate and
their technical assistance. some prototype antiepileptic drugs in mice and rats. Epilepsia 34:11411146.
Olanow CW and Finn AL (1981) Phenytoin: Pharmacokinetics and clinical thera-
References peutics. Neurosurgery 8:112117.
Popoli P, Pezzola A and Sagratella S (1994) Diphenylhydantoin potentiates the EEG
Bauer LA and Blouin RA (1983) Phenytoin Michaelis-Menten pharmacokinetics in
and behavioural effects induced by N-methyl-D-aspartate antagonists in rats.
Caucasian pediatric patients. Clin Pharmacokinet 8:545549.
Psychopharmacology (Berl) 113:471 475.
Danhof M Mandema JW and Stijnen AM (1992a) Pharmacokinetic complexities of
pharmacodynamic studies in vivo, in The in Vivo Study of Drug Action (van Boxtel Rambeck B, Boenick HE, Dunlop A, Mullen PW, Wadsworth J and Richens A (1979)
CJ, Holford NHG and Danhof M eds) pp 3159, Elsevier Science Publishers B. V., Predicting phenytoin dose: A revised normogram. Ther Drug Monit 1:325333.
Amsterdam. Ramsay RE, Hammond EJ, Perchalski RJ and Wilder BJ (1979) Brain uptake of
Danhof M, Voskuyl RA and Hoogerkamp A (1992b) Pharmacokinetic-pharmacody- phenytoin, phenobarbital and diazepam. Arch Neurol 36:535539.
namic modelling with anticonvulsants in animal models and in humans, in The in Ratnaraj N, Goldberg VD and Hjelm M (1989) A micromethod for the estimation of
Vivo Study of Drug Action (van Boxtel CJ, Holford NHG and Danhof M eds) pp free levels of anticonvulsant drugs in serum. Clin Biochem 22:443 450.
179 201, Elsevier Science Publishers B. V., Amsterdam. Ritschel WA and Hussain A (1984) Review on correlation between pharmacological
Danhof M, Mandema JW, Hoogerkamp A and Mathot RA (1993) Pharmacokinetic- response and drug disposition. Methods Find Exp Clin Pharmacol 6:627 640.