Bioavailability and anticonvulsant activity of 2- cyanoguanidinophenytoin, a structural analogue of phenytoin

Article abstract of DOI:10.1021/js960093b

Phenytoin is extensively used in Europe and the United States for the treatment of generalized tonic clonic seizures (grand mal). However, the efficacy is lowered by the erratic bioavailability after oral administration. The current study was conducted in

Full text of DOI:10.1021/js960093b

+
+
Bioavailability and Anticonvulsant Activity of 2-Cyanoguanidinophenytoin, a
Structural Analogue of Phenytoin
X
‡
D
IDIER M. LAMBERT* , BERNARD
M
ASEREEL†, BERNARD
GALLEZ*, MURIEL GEURTS*, AND GERHARD K. E. SCRIBA
Received February 21, 1996, from the *Laboratory of Pharmaceutical Chemistry, Department of Pharmaceutical Sciences, Catholic
University of Louvain, CMFA 73.40, Avenue E. Mounier 73, B-1200 Brussels, Belgium, †Laboratory of Pharmaceutical Chemistry, Institute
of Pharmacy, University of Lie`ge, Rue Fusch 3, 4000 Lie`ge, Belgium, and ‡Department of Pharmaceutical Chemistry, University of
Mu¨nster, Hittorfstrasse 58-62, 48149 Mu¨nster, Germany.
Final revised manuscript received June 4, 1996.
Accepted for
publication July 11, 1996X.
Abstract
0 Phenytoin is extensively used in Europe and the United States
for the treatment of generalized tonic clonic seizures (grand mal).
However, the efficacy is lowered by the erratic bioavailability after oral
administration. The current study was conducted in order to investigate
the physicochemical properties, the bioavailability, and the anticonvulsant
activity of cyanoguanidinophenytoin (CNG-DPH), a structural analogue
of phenytoin, which was obtained by the replacement of the urea moiety
by a cyanoguanidine moiety. CNG-DPH was prepared under homoge-
neous Biltz conditions and under heterogeneous phase-transfer conditions.
CNG-DPH is poorly water soluble and has a pKa of 5.3. At pH 7.4, log
Figure 1sStructures of the compounds.
1,2,5-thiazolidin-3-one 1,1-dioxide,13 and 5,5-diphenyl-2-
iminohydantoin14 have been prepared but showed only a low
anticonvulsant activity in animal models if any activity was
noted at all. However, several 1-carbobenzoxy derivatives of
5,5-diphenyl-2-iminohydantoin showed good activity in the
MES test.14 These observations led us to investigate if the
introduction of an electron-withdrawing cyano group could
restore the anticonvulsant activity and, moreover, improve the
bioavailability of the compound in comparison to phenytoin.
The target compound, 2-(cyanoimino)-5,5-diphenyl-4-imi-
dazolidinone (2-cyanoguanidinophenytoin, CNG-DPH, Figure
1), has been synthesized previously,15,16 but no information
about the pharmacology or the pharmacokinetics is available
in the literature. Thus, the present study was conducted in
order to determine the physicochemical properties, the phar-
macokinetics, and the anticonvulsant activity of CNG-DPH.
P′ was 1.16, from which a pH-independent log P of 3 can be calculated.
Pharmacokinetic parameters were obtained after oral administration of
CNG-DPH to rats and were compared to those of phenytoin after
administration of an equimolar amount. AUC, tmax, and Cmax were
significantly increased compared to those of phenytoin. The anticonvulsant
profile was similar to the profile of phenytoin. CNG-DPH was active in
the maximal electroshock seizure test, albeit 7-fold less active than
phenytoin. The analogue did not protect animals against convulsions
induced by chemicals such as pentylenetetrazole, picrotoxin, N-methyl-
aspartate, strychnine, and bicuculline. It is concluded that while the
bioisosteric exchange of the urea moiety of the molecule with the
cyanoguanidine moiety dramatically changed the physicochemical and
pharmacokinetic parameters compared to those of phenytoin, the
concomitant change of the affinity toward molecular targets reduced the
pharmacological activity and the therapeutic efficacy of the compound.
Materials and Methods
Ch em istr ysMelting points are uncorrected. NMR spectra were
recorded on a Bruker WM 300 (Bruker, Karlsruhe, Germany) instru-
ment. Chemical shifts are reported in parts per million (δ), using
the solvent DMSO-d6 as internal reference. Elemental analyses were
performed on a Carlo-Erba EA 1108 Analyzer (Carlo Erbe, Milano,
Italy) and are within (0.4% of the theoretical values. Infrared spectra
(IR) were obtained on a Perkin-Elmer 457 spectrometer (Perkin
Elmer, Du¨sseldorf, Germany). Mass spectra were recorded on a
Varian MAT 44S (Varian AG, Bremen, Germany), EI: 70 eV, source
temperature 200 °C.
Introduction
Phenytoin (Figure 1) was introduced as an antiepileptic
agent in 19381 and still remains one of the major antiepileptic
drugs in Europe and the United States for the treatment of
generalized tonic clonic seizures (grand mal). In seizure
animal models phenytoin exhibits a specific anticonvulsant
pattern: the compound is active in the maximal electroshock
seizure (MES) test but inactive in the seizure threshold test
with subcutaneous pentylenetetrazol (scMet test) and other
chemoconvulsant tests.2 At therapeutic concentrations, the
drug interacts with voltage- and frequency-dependent sodium
channels.3 However, as a low-water-soluble and low-lipid-
soluble drug the compound shows erratic bioavailability after
oral administration.4,5
Syn th esis of 2-(Cya n oim in o)-5,5-d ip h en yl-4-im id a zolid in on e
(CNG-DP H)sHomogeneous (Biltz) conditionssCNG-DPH was pre-
pared under homogeneous conditions from 19.97 g of benzil (95 mmol),
8.4 g of dicyandiamide (100 mmol), and 10.14 g of potassium hydroxide
(178 mmol) in 325 mL of absolute ethanol as described.15-16 Recrys-
tallization from DMF/water afforded 14 g (50%): mp 254-257 °C dec
(lit.15 mp 260 °C); 13C-NMR (300 MHz, DMSO-d6) 72.30 (C6H5C),
115.27 (CN), 126.99, 128.68, 128.76 (CH arom), 138.92 (C arom),
160.67 (CdNCN), 174.62 (CdO); IR (KBr, cm-1) 2200 (CN), 1755
(CdO), 1640 (CdNCN); MS m/ z ) 276 (M+, 42), 247 (M - 29, 100).
Anal. (C16H12N4O‚H2O) C, H, N.
In order to overcome this drawback, several phenytoin
prodrugs, mostly esters of 3-(hydroxymethyl)phenytoin, have
been investigated.6-9 Fosphenytoin (ACC-9653)10 is currently
under clinical investigation for the treatment of the status
epilepticus.11
Phase-Transfer Conditions17,18sBenzil (19.97 g, 95 mmol), 8.4 g of
dicyandiamide (100 mmol), 10.14 g of potassium hydroxide (178
mmol), and 2.5 g of PEG 600 were heated at 100 °C for 2 h in a two-
phase system consisting of 200 mL of water and 200 mL of 1-butanol.
After cooling, the mixture was filtered and acidified with acetic acid
to give a precipitate, which was washed with water and recrystallized
As an alternative approach, structural analogues of pheny-
toin such as 5,5-diphenyl-2-thiohydantoin,12 4,4-diphenyl-
X Abstract published in Advance ACS Abstracts, September 1, 1996.
© 1996, American Chemical Society and
American Pharmaceutical Association
S0022-3549(96)00093-7 CCC: $12.00
Journal of Pharmaceutical Sciences / 1077
Vol. 85, No. 10, October 1996

+
+
to give 16.5 g (59%). The product gave mp, spectra, and elemental
as the carrier gas at a flow rate of 0.75 mL/min. The column
temperature was maintained at 225 °C. The detector temperature
was set to 300 °C, the injector temperature to 270 °C. The injector
was operated in the splitless mode during the injection and for the
first 10 s of each analysis. A split of 1:10 was applied throughout
the remainder of the analysis. The retention times of phenytoin and
5-(p-methylphenyl)-5-phenylhydantoin were 9.0 and 11.7 min, re-
spectively.
analysis identical with those of the batch synthesized by the homo-
geneous method. Thus, material from both syntheses was pooled and
used for further studies.
Deter m in a tion of th e Aqu eou s Solu bilitysThe solubility of
CNG-DPH was determined by vigorous stirring of a saturated solution
of CNG-DPH at 25 ( 0.2 °C in water or at 37 ( 0.2 °C in 0.05 M
phosphate buffer (µ ) 0.5) at pH 6, 6.5, 7.4, and 8 for 18 h. Each
sample was run in quadruplicate. Following filtration, aliquots of
the clear filtrate were analyzed by HPLC using the system described
under Analysis of Plasma Samples. Concentrations were calculated
from a calibration curve obtained by analysis of the pure compound
under identical chromatographic conditions.
Deter m in a tion of th e p Ka sA 2.5 mL portion of a stock solution
of CNG-DPH (25 × 10-5 M) in methanol was transferred to a 25 mL
volumetric flask and brought to a final volume of 25 mL with buffer
(from pH 2.4 to pH 10) containing methanol (from 0 to 2.5 mL), and
the pH was determined. The UV absorption was recorded from 380
to 210 nm and the absorbance measured at 236 nm. The pKa was
calculated according the following equation:
CNG-DPH and phenytoin concentrations were calculated by the
peak area ratio method using calibration curves obtained from spiked
plasma samples.
Da ta An a lysissNoncompartmental analysis of the data was
performed using TOPFIT 2.0. The maximum plasma concentration
(Cmax) and the time to reach this concentration (tmax) were obtained
directly from the plasma concentration versus time profiles. The area
under the curve (AUC) was calculated employing the log-linear
trapezoidal method for the elimination phase for the observed values
and by extrapolation to infinity. The elimination half-life (t1/2) was
estimated from the final segment of the plasma concentration curve.
P h a r m a cologica l Eva lu a tion sMES TestsMale OF1 mice (If-
faCredo, Les Oncins, France) weighing 18-30 g were used. Corneal
electrodes primed with a drop of sodium chloride solution were applied
to the eyes, and an electrical stimulus (50 Hz, 0.2 s, 50 mA, J ansen
STI stimulator) was delivered. Abolition of the hind-leg tonic extensor
component was noted as protection.19 The compounds were dissolved
in DMSO or suspended in 0.5% aqueous methyl cellulose suspension.
Generally, a volume of 2 mL/kg was applied.
The time of peak effect was determined at 0.5, 1, 2, 3, 4, 6, 8, and
24 h upon intraperitoneal administration of a dose of 109 µmol/kg of
CNG-DPH (30 mg/kg) or 15 µmol/kg of phenytoin (3.8 mg/kg) as
solutions in DMSO using five to eight mice per time point. At the
respective time of peak effects the median effective dose (ED50) was
calculated from at least five different doses (five to eight animals per
dose) according to the method of Litchfield and Wilcoxon.20
Ch em ica l-In d u ced Con vu lsion ssMale OF1 mice (Iffa-Credo,
Les Oncins, France), weighing 18-30 g, received an intraperitoneal
dose of 30, 100, or 300 mg/kg of CNG-DPH or phenytoin at the time
to peak effect of the MES activity of the compounds as recommended
by the ADD program of the NIH.19 At 2 h and at 30 min after the ip
injection of phenytoin and CNG-DPH, respectively, pentylenetetrazole
(85 mg/kg), strychnine (1.2 mg/kg), N-methyl-D,L-aspartate21 (NM-
DLA, 340 mg/kg), bicuculline (2.7 mg/kg), picrotoxin (3.15 mg/kg), or
3-mercaptopropionic acid22 (40 mg/kg) (all from Sigma, Bornem,
Belgium) was administered subcutaneously to five mice per dose and
per convulsant treatment. Each animal was isolated in an individual
cage and observed for 30 min. Clonic and tonic convulsions were
noted.19
pKa ) pH + log[(A- - B)/(B - HA)]
where A- ) absorbance of CNG-DPH at pH 10, B ) absorbance of
CNG-DPH in the buffered solution, and HA ) absorbance of CNG-
DPH at pH 2.4.
Deter m in a tion of log P ′sTwo 5 × 10-5 M solutions of CNG-DPH,
one in 1-octanol and the other in phosphate buffer, pH 7.4, were
prepared. A 10 mL portion of the octanol solution was shaken with
10 mL of buffer for 1 h, and 10 mL of the buffer solution was shaken
with 10 mL of 1-octanol for 1 h. The samples were prepared in
triplicate. Following centrifugation, the UV absorbance of the aque-
ous and organic layers was measured at 236 nm. The log P′ was
calculated according to the following equation:
log P′ ) log{[CNG-DPH](octanol)/[CNG-DPH](buffer)}
P h a r m a cok in eticssMale Wistar rats (bred at the animal facili-
ties at UCL), weighing 300-350 g, were housed individually with a
12 h light-dark cycle and free access to commercial rodent chow and
water. The animals were fasted overnight and during the first 24 h
of the experiment but were allowed water ad libitum. Doses of 119
µmol/kg of CNG-DPH and phenytoin were administered as suspen-
sions in 0.5% methyl cellulose by oral intubation in a volume of 2
mL/kg to three animals (CNG-DPH) and four animals (phenytoin),
respectively. Approximately 300 µL of blood was collected via the
tail clip method into Eppendorf tubes containing 30 µL of a 0.5%
sodium citrate solution. The samples were centrifuged at 4 °C at
2500g for 10 min, and the plasma was immediately separated, frozen
at -80 °C, and stored frozen until analyzed.
An a lysis of P la sm a Sa m p lessPlasma amounts of 100-150 µL
were placed in 10 mL stoppered centrifuge tubes, acidified with 50
µL of 3 M NaH2PO4, pH 3, and extracted by vortexing with 2 mL of
toluene containing 1 µg of 5-(p-methylphenyl)-5-phenylhydantoin as
internal standard. After centrifugation at 2500g for 10 min at room
temperature the organic layer was transferred to a clean centrifuge
Acu te Neu r otoxicitysAt the time of the peak effect of the MES
activity, mice were subjected to the rotorod test19 defining toxicity as
the failure of the treated animal to maintain equilibrium for 1 min
in three trials on a rod rotating at 6 rpm. The median toxic dose
(TD50) was calculated from six doses (seven animals per dose) after
intraperitoneal administration of CNG-DPH (between 10 and 100 mg/
kg) or phenytoin (between 1 and 75 mg/kg) as a solution in DMSO.
Results
tube. CNG-DPH samples were dried under
a gentle stream of
nitrogen. The residue was dissolved in 80 µL of water/acetonitrile
(1:1, v/v), and 20 µL was analyzed by HPLC consisting of a Shimadzu
LC-6A solvent delivery module and a Shimadzu SPD-6A UV detector
(Shimadzu, Duisburg, Germany) set at 230 nm. The separation of
the compounds was obtained on a LiChrosorb RP-18 column (5 µm,
125 × 4.6 mm). The mobile phase consisted of a 0.05 M phosphate
buffer, pH 5.5, containing 25% acetonitrile. The flow rate was 1.5
mL/min. The retention times of CNG-DPH and 5-(p-methylphenyl)-
5-phenylhydantoin were 8.5 and 20.8 min, respectively.
Analysis of the phenytoin samples was conducted as described
above except that the toluene layer was extracted with 25 µL of 0.01
M phenyltrimethylammonium hydroxide/0.1 M tetramethylammo-
nium hydroxide in methanol/water (1:1, v/v) by vortexing for 1 min.
Following centrifugation at 2500g for 5 min at room temperature the
toluene phase was discarded and approximately 1 µL of the aqueous
methanol layer was injected into a Shimadzu GC-14A gas chromato-
graph equipped with a flame ionization detector (Shimadzu, Duisburg,
Germany). The separation of the compounds was obtained on a 25
m HP-1 column (Hewlett Packard, Du¨sseldorf, Germany) with helium
Ch em istr ysThe synthesis of CNG-DPH was achieved
using procedures described for the synthesis of phenytoin. Use
of heterogeneous conditions known to reduce the formation
of the diphenylglycoloril side products16,17 only marginally
improved the yield.
P h ysicoch em ica l P r op er tiessThe physicochemical prop-
erties of CNG-DPH in comparison to those of phenytoin have
been summarized in Table 1. At 25 °C the compound is even
less soluble in pure water than phenytoin. At 37 °C in
buffered solutions, the solubility of CNG-DPH at pH 6 was
0.010 ( 0.001 mg/mL (0.034 ( 0.003 µmol/mL); at pH 6.5,
0.067 ( 0.006 mg/mL (0.23 ( 0.02 µmol/mL); at pH 7.4, 0.126
( 0.011 mg/mL (0.43 ( 0.04 µmol/mL); and at pH 8, 0.57 (
0.23 mg/mL (1.94 ( 0.08 µmol/mL). The pKa was determined
spectrophotometrically. By adding 10%, 15%, and 20% metha-
nol to the buffers in order to dissolve CNG-DPH, pKa values
of 5.48 ( 0.04, 5.55 ( 0.01, and 5.45 ( 0.06, respectively, were
1078 / Journal of Pharmaceutical Sciences
Vol. 85, No. 10, October 1996

+
+
Table 1
s
Physicochemical Properties of Phenytoin and CNG-DPH
Table 2sPharmacokinetic Parameters (Mean ± SD) Obtained from
Plasma Levels after Oral Administration of an Equimolar Dose of 119
Phenytoin
CNG-DPH
0.016
0.002b
5. 49
1.16
µ
mol/kg of Phenytoin (n
)
4) and CNG-DPH (n ) 3) to Rats
Aqueous solubility
pKa
log P
0.03a
8.3c
±
AUC
‚
Cmax
(nmol/mL)
±
0.03b
Compound
(nmol
h/mL)
tmax (h)
t1 2 (h)
/
′
2.4d
Phenytoin
CNG-DPH
40.5
257.9
±
±
6.9
19.1
7.52
18.57
±
±
0.45
2.00
1.25
6.67
±
±
0.5
2.31
3.48
9.81
±
±
0.46
1.54
a Literature values vary from 0.01 to 0.04 mg/mL.7,29,30 b Mean
−
4). c From ref 31. d From ref 32.
± SEM (n )
3
Table 3sAnticonvulsant Activity of Phenytoin and CNG-DPH after
Intraperitoneal Administration to Micea
Phenytoin
CNG-DPH
Time to peak effect
ED50 (MES test)b
TD50 (rotorod test)b
Protective index
2 h
12.0 (7.0
73.2 (22.5
6.1
30 min
90.6 (54.3
−
−
20.8)
195)
µ
mol/kg
mol/kg
−
293)
145)
µ
mol/kg
mol/kg
µ
220 (150
−
µ
2.4
(ED50/TD50
)
Pentylenetetrazol
Strychnine
NMDLA
Bicuculline
Picrotoxin
Inactivec
Inactivec
Inactivec
Inactivec
Inactivec
Inactivec
Inactivec
Inactivec
Inactivec
Inactivec
Inactivec
Inactivec
3-Mercaptopropionic
acid
a The compounds were injected as solutions in DMSO 2 h and 30 min,
respectively, prior to the tests. Chemoconvulsants were administered subcutane-
ously. b Calculated from five to six doses with five to eight animals per dose
according to Litchfield and Wilcoxon;20 95% confidence intervals are given in
parentheses. c Inactive up to 300 mg/kg (corresponding to 1.19 mmol/kg of
phenytoin and 1.02 mmol/kg of CNG-DPH).
Figure 2
s
Mean (
)
±SD) plasma concentrations of phenytoin (b, n ) 4) and CNG-
DPH ( , n
O
3) versus time profile following oral administration of equimolar
doses of the compounds to rats (119
phenytoin and 35 mg/kg of CNG-DPH).
µmol/kg corresponding to 30 mg/kg of
found. Therefore, the pKa of 5.49 ( 0.03 as the mean of the
values can be assumed to be independent of the methanol
concentration. Thus, CNG-DPH is approximately 3 orders of
magnitude more acidic than phenytoin.
The log P′ (Table 1) was measured by the shake flask
method with a UV dosage at 236 nm. A pH-independent log
P can be calculated from the respective log P′ and pKa values
according to the equation
log P ) log P′ + log[1+10(pH-pK )
]
a
Values of log P of 3.03 and 2.45 were calculated for CNG-
DPH and phenytoin, respectively.
P h a r m a cok in eticssPlasma concentrations of CNG-DPH
were determined using a reversed-phase HPLC assay, while
phenytoin plasma samples were analyzed by gas chromatog-
raphy as described.23,24 The HPLC assay for CNG-DPH was
linear in the range of 0.5 and 25 µg/mL (corresponding to 1.7
and 85 nmol/mL). The precision was less than 6% below 2
µg/mL and better than 2% at the other concentrations used.
The detection limit was approximately 0.2 µg/mL (0.7 nmol/
mL). The phenytoin GC assay was linear in the range of 0.5
and 25 µg/mL (corresponding to 2 and 99 nmol/mL) with a
precision of 7.1% at 0.5 µg/mL and better than 2.6% at the
other concentrations used. The detection limit was about 0.2
µg/mL phenytoin (0.8 nmol/mL). The plasma concentration
versus time profiles following oral administration of equimolar
amounts of CNG-DPH and phenytoin, respectively, are shown
in Figure 2. The pharmacokinetic parameters obtained by
noncompartmental analysis of the plasma data are sum-
marized in Table 2. Compared to the application of phenytoin,
administration of CNG-DPH led to higher plasma levels. An
approximate 2.5-fold increase of Cmax, an 8-fold increase of
the AUC, and a 5-fold increase of tmax were observed. The
elimination half-life approximately tripled.
Figure 3
a dose of 103
DMSO; ( ) oral administration as a suspension in 0.5% aqueous methyl cellulose.
The data are expressed as percent animals protected versus total number of
animals tested (n 9 animals per time point).
s
Time course of the MES activity of CNG-DPH after administration of
µ
mol/kg (30 mg/kg) to mice: ( ) ip administration as a solution in
0
O
) 5−
P h a r m a cologica l ActivitysCNG-DPH exhibited an an-
ticonvulsant profile similar to the profile of phenytoin after
intraperitoneal administration to mice. The compound was
active in the MES test, but did not protect the animals from
convulsions induced by pentylenetetrazole, strychnine, picro-
toxin, bicuculline, mercaptopropionic acid, or NMDLA at doses
up to 300 mg/kg (Table 3). The time course of the anticon-
vulsant activity of CNG-DPH determined in the MES test
after ip and oral administration to mice is shown in Figure 3.
The maximal MES activity was observed 30 min and 4 h
following intraperitoneal injection and oral administration,
Journal of Pharmaceutical Sciences / 1079
Vol. 85, No. 10, October 1996

+
+
respectively. The anticonvulsant effect of phenyoin upon ip
CNG-DPH displayed lower toxicity than phenytoin in the
rotorod test. However, due to the concomitant decrease of the
antiepileptic efficacy a protective index of 2.4 was found for
CNG-DPH. The protective index of 6.1 determined for pheny-
toin upon ip administration in DMSO is in the same range as
the index of 6.9 reported in the literature after ip injection of
a suspension in methyl cellulose.19,25 Mice receiving even high
doses of CNG-DPH were well awake and kept their explor-
atory behavior. In contrast, animals treated with similar
doses of phenytoin were heavily sedated. Thus, these results,
combined with the pharmacokinetics, suggest that CNG-DPH
might be a suitable candidate for the exploration of the
peripheral effects of phenytoin such as the antiarrhythmic
action in cardiac muscle tissue.
From theoretical calculations Weaver derived that the -N3-
(H)-C4(dO)-C5-C6H5 moiety of phenytoin defines the bio-
active fragment responsible for the anticonvulsant activity
while the -C2(dO)-N3(H)- moiety represents the biotoxic
fragment responsible for DNA interaction.26 However, 2-des-
oxyphenytoin which was suggested to be an improved pheny-
toin analogue,26 is less active in the MES test than the parent
compound.27 A lower anticonvulsant activity in the MES test
has also been reported for 2-thiophenytoin12 and has been
found for CNG-DPH in this study. In addition, X-ray struc-
ture determinations revealed similar conformations between
CNG-DPH and phenytoin.28 Therefore, in contrast to Weav-
er’s hypothesis, the present results in accordance with litera-
ture data suggest that the C-2 carbonyl is essential for optimal
binding of the hydantoin derivatives to the molecular targets.
It is concluded, while the bioisosteric exchange of the urea
moiety with the cyanoguanidino moiety yielding CNG-DPH
dramatically changed physicochemical and pharmacokinetic
parameters compared to those of phenytoin, the concomitant
change of the affinity toward molecular targets reduced the
pharmacological activity and, thus, the therapeutic efficacy
of the compound. Currently, further 5,5-diphenyl-2-imino-
hydantoin derivatives with electron-withdrawing groups in
position 2 are synthesized for the evaluation of their anticon-
vulsant efficacy and the antiarrythmic action in cardiac
muscle.
administration in DMSO peaked after 2 h (data not shown).
On the basis of the median effective dose (ED50) obtained after
ip application, CNG-DPH was about 7.5-fold less active than
phenytoin. CNG-DPH exhibited decreased toxicity in the
rotorod test (Table 3) but a lower protective index (TD50/ED50)
compared to phenytoin. Thus, indices of 6.1 and 2.4 were
calculated for phenytoin and CNG-DPH, respectively. The
vehicle DMSO did not exhibit any activity in the MES test or
the rotorod test at a dose of 2 mL/kg.
Discussion
CNG-DPH was synthesized as a bioisostere of phenytoin,
replacing the urea moiety by the cyanoguanidine group. In
the development of several drugs such as cimetidine, a
analogous bioisosteric substitution has been successful. CNG-
DPH was synthesized in 1970,14 but no detailed study on the
pharmacology or the pharmacokinetics has been performed.
CNG-DPH was synthesized from benzil and dicyandiamide
under homogeneous Biltz conditions and under heterogeneous
conditions in 1-butanol and water using PEG 600 as phase-
transfer catalyst. The latter system is known to reduce the
formation of diphenylglycoloril side products,17,18 thus increas-
ing the yield. However, in the present synthesis only a small
improvement of the yield of CNG-DPH was noted when
heterogeneous conditions were used.
CNG-DPH plasma concentrations were analyzed by HPLC
while phenytoin plasma levels were measured by a GC
procedure developed previously.23 In general, CNG-DPH can
also be analyzed by the GC assay of phenytoin employing flash
methylation with phenyltrimethylammonium hydroxide and
tetramethylammonium hydroxide. Under these conditions
CNG-DPH is converted to 1,3-dimethylphenytoin (G.K.E.S.
and D.M.L., in preparation). However, this procedure is less
sensitive than the HPLC assay probably due to incomplete
conversion in the injector. Thus, different assays were used
for the determination of CNG-DPH and phenytoin plasma
levels.
Compared to phenytoin, CNG-DPH exhibited a significantly
increased Cmax and bioavailability upon oral administration.
Although the compound is poorly water soluble at pH < 6,
the solubility gradually increases at higher pH, in agreement
with the pKa value of 5.49. Thus, as the pH increases along
the intestines, CNG-DPH is dissolved and absorbed. The pH
dependency of the aqueous solubility of CNG-DPH also
explains the delayed tmax. Although CNG-DPH is predomi-
nantly ionized at pH 7.4, the lipophilicity reflected in a log P′
value of 1.16 appeared to ensure sufficient penetration of the
intestinal barrier.
In mice, CNG-DPH exhibits an anticonvulsant profile
similar to that of phenytoin. Both compounds were active in
the MES test but inactive in chemical-induced convulsion
models such as subcutaneously injected pentylenetetrazole,
strychnine, NMDLA, bicuculline, picrotoxin, or 3-mercapto-
propionic acid. This pattern is typical for phenytoin and
consistent with literature data.19,25 J udged from the anticon-
vulsant profiles it might be speculated that CNG-DPH
interacts with similar molecular targets as phenytoin, i.e.,
sodium channels. The mechanism of action does not seem to
implicate neurotransmitter systems such as GABA, glycine,
and glutamate because no activity could be detected in the
chemoconvulsion tests. On the basis of the ED50 values upon
ip administration, CNG-DPH is approximately 7-fold less
active than phenytoin. This decrease in the anticonvulsant
activity in comparison to phenytoin might be attributed to a
lower affinity toward the molecular targets such as sodium
channels and/or to a reduced penetration of the blood-brain
barrier due to the higher degree of ionization.
References and Notes
1. Merrit, H. H.; Putnam, T. J . J . Am. Med. Assoc. 1978, 111,
1068-1073.
2. Krall, R. L.; Penry, J . K.; White B. G.; Kupferberg, H. J .;
Swinyard, E. A. Epilepsia 1978, 19, 404-428.
3. MacDonald, R. L.; Kelly, K. M. Epilepsia 1995, 36, S2-S12.
4. Suzuki, T.; Saitoh, Y; Nishihara, K. Chem. Pharm. Bull. 1970,
18, 405-411.
5. Arnold, K.; Gerber, N.; Levy, G. Can. J . Pharm Sci. 1970, 5,
89-92.
6. Shek, E. Adv. Drug Delivery Rev. 1994, 14, 227-241.
7. Varia, S. A.; Schuller, S.; Sloan, K. B.; Stella, V. J . J . Pharm.
Sci. 1984, 73, 1068-1073.
8. Yamaoka, Y.; Roberts, R. D.; Stella, V. J . J . Pharm. Sci. 1983,
72, 400-405.
9. Scriba, G. K. E. Arch. Pharm. 1993, 326, 477-481.
10. Smith, R. D.; Brown B. S.; Maher, R. W.; Matier W. L. Epilepsia
1989, 30, S15-S21.
11. Bebin, B.; Bleck, T. P. Drugs 1994, 48, 153-171.
12. Sohn, Y. I.; Levitt, B.; Raines, A. Arch. Int. Pharmacodyn. 1970,
188, 284-289.
13. Lee, C. H.; Kohn, H. J . Pharm. Sci. 1990, 79, 716-718.
14. Kwon, C.-H.; Iqbal, M. T.; Wurpel, J . N. D. J . Med. Chem. 1991,
34, 1845-1849.
15. Call, L. Monatsh. Chem. 1970, 101, 344-346.
16. Schramm, H. W. Sci. Pharm. 1989, 57, 385-390.
17. Mergen, F.; Poupaert, J . H.; De Keyser, J . L.; Dumont P.
Pharmazie 1989, 44, 110-112.
18. Poupaert, J . H.; De Keyser, J . L.; Vandervorst, D.; Dumont, P.
Bull. Soc. Chim. Belg. 1984, 93, 493-495.
1080 / Journal of Pharmaceutical Sciences
Vol. 85, No. 10, October 1996

+
+
19. Porter, R. J .; Cerghino, J . J .; Gladding, G. D.; Hessie, B. J .;
Kupferberg, H. J .; Scoville, B.; White, B. G. Cleveland Clin. Q.
1984, 51, 293-305.
27. Poupaert, J . H.; Vandervorst, D.; Guiot, P.; Moustafa, M. M.
M.; Dumont, P. J . Med. Chem. 1984, 27, 76-78.
28. Dupont, L.; Masereel, B.; Lambert, D. M.; Scriba, G. K. E. Acta
Crystallogr. 1995, C51, 1901-1903.
20. Litchfield, J . T.; Wilcoxon, F. J . Pharm. Exp. Ther. 1949, 96,
99-102.
29. Bundgaard, H.; J ohansen, M. Int. J . Pharm. 1980, 5, 67-77.
30. Schwartz, P. A.; Rhodes, T. C.; Cooper, J . W. J r. J . Pharm. Sci.
1977, 66, 994-997.
31. Agarwal, S. J .; Blake, M. I. J . Pharm. Sci. 1968, 57, 1434-1435.
32. Le Petit, G. Pharmazie 1977, 32, 289-291.
21. Czuczwar, S. J .; Frey, H. H.; Lo¨scher, W. Eur. J . Pharmacol.
1985, 108, 273-280.
22. Seiler, N.; Sarhan, S. Naunyn-Schmiedeberg’s Arch. Pharmacol.
1984, 326, 49-57.
23. Scriba, G. K. E.; Lambert, D. M.; Poupaert, J . H. J . Pharm. Sci.
1995, 84, 300-302.
Acknowledgments
B.G. and M.G. are Senior Research Assistant and Research
Assistant, respectively, of the Belgian National Fund for Scientific
Research.
24. Stella, V. J . J . Pharm. Sci. 1977, 66, 1510-1511.
25. Meldrum, B. S. Preclinical test systems for evaluation of novel
compounds in Current problems in epilepsy 4. New Anticonvul-
sant drugs; Meldrum, B. S., Porter, R. J ., Eds.; J ohn Libey:
London, 1986; pp 31-48.
26. Weaver, D. F. Seizure 1992, 1, 223-240.
J S960093B
Journal of Pharmaceutical Sciences / 1081
Vol. 85, No. 10, October 1996