Full text of DOI:10.1038/sj.bjp.0703337

British Journal of Phar
m
acology (2000) 130, 242 ± 248
ã
2000 Macmillan Publishers Ltd All rights reserved 0007 ± 1188/00 $15.00
www.nature.com/bjp
Comparison of serum, cerebrospinal ¯uid and brain extracellular
¯uid pharmacokinetics of lamotrigine
*,1M.C. Walker, 1X. Tong, 1H. Perry, 1M.S. Alavijeh & 1P.N. Patsalos
1Epilepsy Research Group, Pharmacology and Therapeutics Unit, University Department of Clinical Neurology, Institute of
Neurology, Queen Square, London WC1N 3BG
1
We investigated the rate of penetration into and the intra-relationship between the serum,
cerebrospinal ¯uid (CSF) and regional brain extracellular ¯uid (bECF) compartments following
systemic administration of lamotrigine in rat.
2
The serum pharmacokinetics were biphasic with an initial distribution phase, (half-life
approximately 3 h), and then prolonged elimination phase of over 30 h. The serum
pharmacokinetics were linear over the range 10 ± 40 mg kg71
Using direct sampling of CSF with concomitant serum sampling, the calculated penetration half-
a
.
3
time into CSF was 0.42+0.15 h. At equilibrium, the CSF to total serum concentration ratio
(0.61+0.02) was greater than the free to total serum concentration (0.39+0.01).
4
Using in vivo recovery corrected microdialysis sampling in frontal cortex and hippocampus with
concomitant serum sampling, the calculated penetration half-time of lamotrigine into bECF,
0.51+0.11 h, was similar to that for CSF and was not area or dose dependent. At equilibrium, the
bECF to total serum concentration ratio (0.40+0.04) was similar to the free to total serum
concentration (0.39+0.01), and did not di€er between hippocampus and frontal cortex.
5
The species speci®c serum kinetics can explain the prolonged action of lamotrigine in rat seizure
models. Lamotrigine has a relatively slow penetration into both CSF and bECF compartments
compared with antiepileptic drugs used in acute seizures. Furthermore, the free serum drug
concentration is not the sole contributor to the CSF compartment, and the CSF concentration is an
overestimate of the bECF concentration of lamotrigine.
British Journal of Pharmacology (2000) 130, 242 ± 248
Keywords: Lamotrigine; cerebrospinal ¯uid; pharmacokinetics; brain extracellular ¯uid; microdialysis; antiepileptic drug
Abbreviations: AED, antiepileptic drug; bECF, brain extracellular ¯uid; Cmax, maximum concentration; CSF, cerebrospinal
¯uid; Ct, concentration at time t; FC, frontal cortex; HC, hippocampus; h.p.l.c., high performance liquid
chromatography; LTG, lamotrigine; Tmax, time to maximum concentration
Introduction
Lamotrigine
(3,5-diamino-6-(2,3-dichlorophenyl)-1,2,4-tria-
such drugs is critical for: optimization of therapy, determining
the value of serum concentration monitoring, and ascertaining
drug modes and mechanisms of action. Additionally it is not
enough to picture the brain as a single compartment. The brain
consists of extracellular, intracellular, and cerebrospinal ¯uid
(CSF) compartments and depending on where a drug acts
depends on which compartment's kinetics is of most relevance.
The constitution of and access to these compartments is
di€erent (Davson & Segal, 1995). The pharmacokinetics of
AEDs in whole brain are determined by non-speci®c binding
to brain lipids and proteins; they are thus unlikely to represent
the pharmacodynamically relevant compartment. Receptors
on neurons, and ion channels on axons are surrounded by
brain extracellular ¯uid (bECF), and it is likely that the
pharmacodynamics of drugs that act on these receptors and
ion channels are determined by the unbound concentration of
drugs in the bECF (Sechi et al., 1989). Although, the bECF
and CSF are produced independently, they are in direct
communication with one another so that changes in the
composition of one are often re¯ected in changes in the
composition of the other (Davson & Segal, 1995). CSF drug
concentrations could thus be an indirect index of bECF
concentrations. Measurements of CSF penetration, however,
do not necessarily give an accurate indication of blood brain
permeability, and there are circumstances when CSF con-
centrations are a poor indicator of brain tissue and bECF
concentrations (Thomas & Segal, 1998).
zine) is a phenyltriazine derivative. In clinical trials, lamo-
trigine has been shown to be an e€ective antiepileptic drug
(AED) with a broad spectrum of activity (Fitton & Goa, 1995).
It protects against lesions produced by kainate (McGeer &
Zhu, 1990); at high doses it can reduce cortical infarct volume,
and protect against global cerebral ischaemia (Rataud et al.,
1994; Smith & Meldrum, 1995; Crumrine et al., 1997; Wiard et
al., 1995). It thus has potential to be used acutely as a
neuroprotectant. Its mechanism of action is, in part, mediated
through inhibition of voltage-activated sodium channels, and
also an action on pre-synaptic N-type and P-type calcium
channels (Walker & Sander, 1999). It has a favourable
pharmacokinetic pro®le in humans with good absorption,
linear pharmacokinetics and minimal e€ect on the pharmaco-
kinetics of other drugs (Fitton & Goa, 1995; Walker & Sander,
1999).
Although serum concentration monitoring of AEDs is
widely and e€ectively used in the management of patients, its
primary purpose is as an index of brain concentrations (i.e. the
concentration at the site of action). Serum pharmacokinetics
may, however, be a very poor index of brain pharmacokinetics
following acute administration of AEDs such as are used in the
treatment of status epilepticus (Sechi et al., 1989; Wilder et al.,
1977). An understanding of the neuropharmacokinetics of
*Author for correspondence; E-mail: mwalker@ion.ucl.ac.uk

M.C. Walker et al
Central and peripheral kinetics of lamotrigine in rat
243
In order to study CSF pharmacokinetics, direct sampling of
CSF with simultaneous serum sampling has been successfully
used (Patsalos et al., 1992; Semba et al., 1993; Lolin et al.,
1994; Walker et al., 1998; Doheny et al., 1999; Nagaki et al.,
1999). The limited accessibility of CSF and the impracticability
of repeated sampling in humans have meant that most of these
drug studies have been carried out in animal models.
then hourly to 12 h and then 3 hourly to 30 h after lamotrigine
administration. After each sampling, the catheter was ¯ushed
with 100 ml of 5 u ml71 heparinized saline to maintain patency
and prevent hypovolaemia. CSF samples were taken at 20 ±
30 min intervals for the ®rst hour, then hourly to 12 h and then
3 hourly to 30 h after lamotrigine administration. Sera were
separated by centrifugation, and sera and CSF were stored at
7708C until analysis.
Various methods have been used to study drug pharmaco-
kinetics in bECF; bECF phenytoin has been measured in the
left temperoparietal region using implanted polypropylene
balls in dogs (Sechi et al., 1989). The technique of microdialysis
has enabled the temporal measurement of bECF drug
concentrations in discrete brain areas following peripheral
administration. This technique has a temporal and spatial
resolution suitable for the study of drug neuropharmacoki-
netics, without signi®cant perturbation to the system measured
(the method does not rely upon the extraction of ¯uid from a
compartment of interest). Importantly the blood-brain barrier
is intact shortly following slow implantation of microdialysis
probes (de-Lange et al., 1997). Microdialysis is, thus, highly
suited for studies of drug neuropharmacokinetics (Patsalos et
al., 1995; de-Lange et al., 1997). The further advantage of the
microdialysis technique is that it enables simultaneous
measurement in two or more brain areas (de-Lange et al.,
1995; Walker et al., 1996)
bECF and blood sampling
Concentric dialysis probes with Filtral 12 (Hospal, Rugby,
U.K.) dialysis membrane 4 mm long, 200 mm diameter were
prepared as previously described (Hutson et al., 1985). On the
day of surgery the in vitro recovery for each probe was
calculated by placing the probes in a 8 mM solution of
lamotrigine dissolved in arti®cial CSF at 378C, and then
perfusing the probes with arti®cial CSF at 2 ml min71. Samples
(40 ml) were collected every 20 min for 80 min, and stored at
7708C until analysis.
Rats were anaesthetized with halothane (2%), and
microdialysis probes were implanted, and the jugular vein was
catheterized using previously described procedures. The probes
were slowly implanted in the hippocampus (from bregma
5.6 mm posterior, 5 mm lateral, 8.2 mm ventral) and the
frontal cortex (from bregma 2.5 mm anterior, 1.5 mm lateral,
5.5 mm ventral) according to the atlas of Paxinos & Watson
(1986), and held in place with dental cement and anchor screws
(De Trey, Surrey, U.K.).
Using well-characterized freely behaving rat models that
permit simultaneous sampling of CSF and blood or bECF and
blood, we set out to characterize the temporal pharmacoki-
netic and neuropharmacokinetic inter-relationship of lamo-
trigine after acute administration. The contribution of dose
and protein binding were also determined.
Two days after surgery when the animals were fully
recovered (Patsalos et al., 1992), the jugular vein catheter,
and dialysis probes were checked for patency. Arti®cial CSF
was perfused through the microdialysis probes at 2 ml min71
.
Methods
Three baseline dialysate samples (40 ml) and concomitant
blood samples (200 ml) were taken in the ®rst hour. The rats
were then injected intraperitoneally with 20 or 40 mg kg71
lamotrigine in propylene glycol; 40 mg kg71 was higher than
that used in the CSF experiments, because following
10 mg kg71 i.p. lamotrigine, the concentrations of lamotrigine
in dialysate would have been below the detectable limit of the
h.p.l.c. analysis. Venous blood samples (200 ml) were with-
drawn every 20 ± 120 min and thereafter hourly until 300 min
after lamotrigine administration. Dialysate samples (20 ml)
were collected every 10 min for 120 min and then every 20 min
Drugs
Lamotrigine (Glaxo-Wellcome, Cheshire) was dissolved in
50% propylene glycol (20 mg ml71
administration or at low concentrations in arti®cial CSF
(composition mM: NaCl 125, KCl 2.5, MgCl2 1.18 and CaCl2
1.26) for recovery experiments.
) for intraperitoneal
Animals
(40 ml) for
a further 180 min. Sera were separated by
Male Sprague-Dawley rats (Charles River, Margate, Kent
U.K.) weighing 260 ± 360 g were used. Rats were individually
housed under a 12 h light-dark cycle (lights on 0800 h), an
ambient temperature of 258C and with free access to water and
to a normal laboratory diet (SDS R and M number 1
expanded, Scienti®c Dietary Services, Witham, Essex, U.K.).
centrifugation, and sera and dialysate were stored at 7708C
until analysis.
In vivo microdialysis probe recovery
The no-net-¯ux method for calculating the di€erence between
in vivo and in vitro recoveries was used (Lonnroth et al., 1987).
Concentric dialysis probes with Filtral 12 (Hospal, Rugby,
U.K.) dialysis membrane 4 mm long, 200 mm diameter were
prepared as previously described.
On the day of surgery the in vitro recovery for each probe
was calculated by placing the probes in a 40 mM solution of
lamotrigine dissolved in arti®cial CSF at 378C, and then
perfusing the probes with arti®cial CSF at 2 ml min71. Samples
(40 ml) were collected every 20 min for 80 min, and stored at
7708C until analysis.
Rats were anaesthetized with halothane (2%), and
microdialysis probes were implanted using previously de-
scribed techniques. The probes were slowly implanted in the
frontal cortex (from bregma 2.5 mm anterior, 1.5 mm lateral,
5.5 mm ventral) according to the atlas of Paxinos & Watson
CSF and blood sampling
Rats were anaesthetized with halothane (2%), and catheters
were implanted in the cisterna magna for CSF sampling and
the right jugular vein for blood sampling, using a previously
described technique except that the position of the intersliding
polythene tubing in the cisterna magna catheter was ®xed with
epoxy resin after implantation (Patsalos et al., 1992).
Two days later the CSF and blood catheters were checked
for patency, and blood (200 ml) and CSF (20 ml) samples were
collected at 30 min intervals for 1 h. Thirty minutes later the
animals were given lamotrigine in propylene glycol by
intraperitoneal injection (10 or 20 mg kg71). Venous blood
samples (200 ml) were withdrawn at 10, 20, 30, 40 and 60 min,
British Journal of Pharmacology, vol 130 (2)

244
M.C. Walker et al
Central and peripheral kinetics of lamotrigine in rat
(1986), and held in place with dental cement and anchor screws
(De Trey, Surrey, U.K.).
cases b44a, and indeed it was not possible to calculate b
accurately despite 30 h of sampling. Thus for ease and
accuracy of modelling the equation was approximated to
Ct=Ae7at+B.
One day after surgery, the rats were injected with
lamotrigine, 40 mg kg71 i.p. 20 h later, arti®cial CSF was
perfused through the microdialysis probes at 2 ml min71
.
The dialysate concentrations were corrected for probe
recovery (see above). The ratios of CSF concentration to
serum concentration and the ratios of corrected dialysate
concentration to serum concentration were calculated for each
individual animal at each time point for which there were
concurrent data. The ratio versus time graphs were modelled
using R=Rss(1-e7kt) where R is the ratio of CSF or dialysate
concentration to serum concentration, Rss is the ratio at steady
state and k is the rise constant (Loscher & Frey, 1984; Mayer et
al., 1959). Where appropriate pharmacokinetic parameters
were compared using analysis of variance (ANOVA).
Three baseline dialysate samples (60 ml) were taken in the ®rst
1.5 h. Then 5, 10, 15 and 20 mM lamotrigine in arti®cial CSF
was perfused through the microdialysis probe 2 ml min71 for
0.5 h after which a dialysate sample (60 ml) was collected over
0.5 h. After the lamotrigine perfusions, arti®cial CSF was
perfused through the microdialysis probes at 2 ml min71 for
0.5 h and a further three dialysate samples (60 ml) were
collected. Dialysate was stored at 7708C until analysis.
Lamotrigine analysis
The concentrations of lamotrigine in sera, CSF and dialysate
were determined by high performance liquid chromatography
(h.p.l.c.) with ultraviolet detection as follows. Sera (50 ml) and
acetonitrile (50 ml) containing 3.75 mg ml71 10-methoxycarba-
mazepine as the internal standard were pipetted into a 1.5 ml
polyethylene tube (Tre€ AG, Switzerland), vortex mixed and
then centrifuged for 5 min at 95006g (Abbott Micro-
Centrifuge, Abbott, Maidenhead, U.K.). Fifty ml of the
supernatant extract were mixed with 25 ml of mobile phase
and 10 ml were injected into the h.p.l.c. system. The h.p.l.c.
system comprised a Spectrasystem AS3000 autosampler, a
Spectrasystem P4000 pump, a Spectrasystem UV2000 detector
and an SP 4270 integrator/printer plotter with LABNET data
system controller (Spectra-Physics, Maidenhead, U.K.). For
CSF and dialysate analysis, 20 ml were directly injected into the
h.p.l.c. system without acetonitrile extraction. Chromatograms
were run at ambient temperature on a Merck LiChroCART
column (12564.0 mm) packed with Lichrosorb RP-8, 5 mm
(BDH, Poole, U.K.). Mobile phases of 0.045 M phosphate
bu€er with 15% acetonitrile and 0.0275 M phosphate bu€er
with 37.5% acetonitrile were run on a gradient. The column
e‚uent was monitored at 215 nm with a sensitivity range of
0.02 absorption units full scale. The procedures for determin-
ing the non-protein bound, serum lamotrigine concentration
were exactly that for sera, except that the sera were ®rst ®ltered
through an Amicon Centrifree Micropartition System (Ami-
con, Stonehouse, U.K.) at a temperature of 258C using a
Sorvall RC-5B refrigerated centrifuge (Du Pont, Stevenage,
U.K.), and 50 ml of the ultra®ltrate was used in the analysis.
Results
Serum pharmacokinetics
The serum concentration versus time pro®les of lamotrigine
during prolonged sampling (Figure 1a) demonstrate rapid
absorption following intraperitoneal injection with peak
concentrations achieved at 0.2 ± 1.0 h for 20 mg kg71 and
0.2 ± 0.5 h for 10 mg kg71 (Table 1). Thereafter there is a
biphasic fall in sera concentrations. The initial alpha phase has
a half-life of approximately 3 h at both doses (Table 1). The
beta phase is prolonged and an accurate estimate of the decay
constant was not possible despite 30 h of sampling. The serum
lamotrigine concentration-time pro®le for the rats with CSF
catheters at a dose of 20 mg kg71 was similar to that for rats
with microdialysis probes at the same dose (Figure 1b). Over
5 h of sampling, the serum lamotrigine concentration-time
pro®les were linearly related to dose (Figure 1b); the area
under the curve for the concentration-time pro®les over 5 h
corrected for dose were not signi®cantly di€erent (P40.1).
Over the concentration range of 5 ± 52 mmol l71
, the
lamotrigine non-protein bound/total serum concentration
ratio was 0.39+0.01 (n=27, Figure 1c).
Microdialysis probe recoveries
A comparison of in vivo and in vitro recoveries were made in
three rats, using the ®nding that the serum lamotrigine
concentrations from 15 ± 30 h after administration do not
signi®cantly change. The rats were injected with lamotrigine 1
day after surgery, and the study was performed the next day
(i.e. at the same time interval for the other microdialysis
experiments). The in vitro recovery of these probes was
10.1+0.9%. The di€erence between dialysate concentration
of lamotrigine (lamotrigine out) and perfusate concentration
of lamotrigine (lamotrigine in) for ®ve di€erent perfusate
lamotrigine concentrations was linearly related to perfusate
concentration of lamotrigine (Figure 3a). The calculated in vivo
recovery using the no-net-¯ux method was 4.8+0.3%. To
con®rm that the extracellular concentration of lamotrigine had
not changed over the study period (a necessary prerequisite of
the no-net-¯ux method), we compared the dialysate lamo-
trigine concentration whilst perfusing with arti®cial CSF
containing no lamotrigine at the beginning of the study and
at the end (0.80+0.07 and 0.90+0.03 mM, respectively;
P40.4). The ratio of in vitro to in vivo recoveries was
2.11+0.16, and this ®gure was used subsequently to correct
the in vitro probe recoveries.
Pharmacokinetic analysis and statistics
For determination of in vivo microdialysis probe recoveries, the
no-net-¯ux method was used (Lonnroth et al., 1987). The
di€erence between lamotrigine concentration out and lamo-
trigine concentration in for each perfusate concentration was
plotted against lamotrigine concentration in. At steady-state
concentrations, the slope of the line through the points
represents the dialysis recovery for lamotrigine. A least mean
squares linear regression analysis was used to calculate the
slope. The in vivo recoveries were compared to the in vitro
recoveries in order to calculate an in vivo to in vitro recovery
ratio; this was used to correct all subsequent in vitro recoveries.
For serum pharmacokinetics: time to maximum concentra-
tion (Tmax) and maximum concentration (Cmax) were estimated
from the graphs. The serum data for prolonged sampling
(30 h) were modelled using Microcal Origin v5 assuming a
biexponential decay, Ct=Ae7at+Be7bt where Ct is the
concentration at time t, b represents the elimination rate
constant, and a represents a redistribution rate constant. In all
British Journal of Pharmacology, vol 130 (2)

M.C. Walker et al
Central and peripheral kinetics of lamotrigine in rat
245
CSF and bECF neuropharmacokinetics
Lamotrigine was detectable in the CSF at the ®rst time point,
10 min post-dose (Figure 2a): At both doses the concentration
then rose over a period of 0.5 ± 1 h up to a maximum, and then
slowly declined. Similarly, lamotrigine appears rapidly in brain
bECF (Figure 3b). The neuropharmacokinetics of lamotrigine
in hippocampus and frontal cortex were indistinguishable
(Figure 3b). The CSF to serum concentration ratio increased
to a steady state value of 0.61 with a half-time of 0.42+0.15 h
(Figure 2b), and the corrected dialysate to serum concentration
ratio increased to a steady state value of 0.40 with a half-time
of 0.51+0.11 h (Figure 3c). Using multifactorial ANOVA,
there was no signi®cant e€ect of dose, brain area (hippocam-
pus vs frontal cortex) or ¯uid compartment (CSF or bECF) on
the rise constant. The steady state ratio was not dose
dependent, but was dependent on compartment with a
signi®cant di€erence between CSF and corrected bECF
dialysate to serum concentration ratios (P50.001).
Discussion
Although as part of its preclinical development many studies
were undertaken to determine the blood (serum) pharmacoki-
netics of lamotrigine in a variety of species, there is a sparcity
of published data (Parsons et al., 1995). Furthermore, as far as
we are aware, there are no published data describing acute
neuropharmacokinetics of lamotrigine and this is the ®rst
report on the temporal inter-relationship of lamotrigine serum
Figure 1 Serum pharmacokinetics following intraperitoneal injection
of lamotrigine (LTG: 10, 20 or 40 mg kg71) in (a) rats with CSF
catheters over 30 h and (b) rats with microdialysis probes compared
with serum concentrations from rats with CSF catheters over 5 h.
Data are presented as mean+s.e.mean of 5 ± 6 rats. (c) Serum
lamotrigine free/total concentration ratio at di€ering total lamotri-
gine concentrations demonstrating no concentration e€ect on
lamotrigine serum protein binding. Dotted line indicates mean value
of ratio.
Table 1 Serum pharmacokinetic constants following in-
traperitoneal injection of lamotrigine
t
for alpha
phase (h)
Tmax
(h)
Cmax
(mmol 171
)
‰
Dose
Rat
10 mg kg71
1
2
3
4
5
2.5
6.0
2.0
0.6
6.2
0.6
3.0
1.0
0.5
0.2
1.0
0.2
0.3
0.2
0.4
0.1
25.1
24.5
17.6
22.6
17.5
31.0
23.0
2.1
6
Mean
s.e.m.
20 mg kg71
1
2
3
4
5
1.9
0.8
4.3
1.1
5.7
2.8
0.9
0.7
1.0
0.2
0.5
0.2
0.5
0.2
30.4
26.0
62.1
60.9
53.6
46.6
7.7
Figure 2 (a) Cerebrospinal ¯uid pharmacokinetics of lamotrigine
following intraperitoneal injection of lamotrigine (LTG: 10 or
20 mg kg71). Data are presented as mean+s.e.mean of 5 ± 6 rats.
(b) Lamotrigine cerebrospinal ¯uid:serum concentration ratio-time
pro®le. Dotted line indicates free serum concentration ratio.
Continuous line indicates modelled values (see Methods). Data are
presented as mean+s.e.mean of eight rats.
Mean
s.e.m.
t
C
=half life, Tmax=time to maximum concentration,
max=maximum concentration.
‰
British Journal of Pharmacology, vol 130 (2)

246
M.C. Walker et al
Central and peripheral kinetics of lamotrigine in rat
signi®cantly less than the CSF replacement rate, and would
thus not be expected and has not been observed to perturb
signi®cantly the concentration pro®le of other antiepileptic
drugs (Patsalos et al., 1992; Lolin et al., 1994; Nagaki et al.,
1999).
After intraperitoneal administration, lamotrigine rapidly
appeared in serum (mean Tmax 0.5 h) suggesting ready
penetration from the peritoneal cavity. The serum pharmaco-
kinetics are biphasic with the ®rst phase probably representing
distribution from the blood compartment, and the second
phase representing mainly elimination. The prolonged elimina-
tion phase in rat is due to rat's inability to eciently conjugate
lamotrigine, and indeed, the method of elimination di€ers
signi®cantly between rat and man (Dickins et al., 1995). In
addition, lamotrigine accumulates in the kidney of male rats
such that the kidney to serum concentration ratio can be up to
300 : 1 (Parsons et al., 1995). This e€ect is both gender and
species speci®c, and could also partly explain the maintenance
of serum concentrations in our study. These kinetic e€ects
would predict a prolonged action of lamotrigine in rat models
of epilepsy, and are important in the determination of
lamotrigine's e€ects in chronic animal models (e.g. kindling),
and in comparing lamotrigine with other AEDs in rat models.
Indeed, in a comparison of AEDs in the maximal electroshock
model, oral lamotrigine had the longest duration of action of a
number of AEDs; it had an almost constant ED50 from 1 ± 8 h
post dose, and at 25 h post dose the ED50 was only 3 ± 4 times
the peak ED50 (Miller et al., 1986).
The mean+s.e.mean serum free fraction of lamotrigine as
determined by ultracentrifugation was 0.39+0.01% and was
not concentration dependent over the concentration range of
5 ± 52 mmol l71. This is similar to a previously reported value
of 0.46 for rat serum determined by equilibrium dialysis
(Parsons et al., 1995).
In order to determine the rate of entry into the CSF and
bECF compartment, we used a mathematical model that
assumes that the rise in serum concentration is instantaneous
and that the serum concentration then does not fall. It remains
a useful estimate in situations in which there is a rapid rise in
serum concentration and then a slow decay, as with our data
(Loscher & Frey, 1984; Mayer et al., 1959). The e€ect of using
intraperitoneal injection and so a slower rise time for the serum
concentration is to underestimate the half-time for rise of the
CSF or bECF to concentration ratio. Thus the estimated
penetration half-times of 0.42 h for CSF and 0.51 h for bECF
are likely to be underestimates. Lamotrigine thus has a slow
rate of CNS penetration, which is slower than values derived
for phenytoin, phenobarbitone, and the benzodiazepines
(Loscher & Frey, 1984). Indeed, this slow penetrance could
militate against lamotrigine's use in acute seizures.
Figure 3 Extracellular ¯uid pharmacokinetics of lamotrigine. (a)
Calculation of in vivo microdialysis probe recoveries. Cout is the
concentration of lamotrigine in dialysate. Cin is concentration of
lamotrigine in perfusate. Slope of line through points gives recovery
(70.048+0.003), and intercept gives extracellular concentration
(715.6+1.3 mmol l71). Data are presented as mean+s.e.mean for
three rats. (b) Microdialysis concentrations of lamotrigine corrected
for in vivo recovery in hippocampus (HP) and frontal cortex (FC)
following intraperitoneal injection of lamotrigine (LTG: 20 or
40 mg kg71). Data are presented as mean+s.e.mean of six rats. (c)
Lamotrigine bECF:serum concentration ratio-time pro®le. Dotted
line indicates free serum concentration ratio. Continuous line
indicates modelled values (see Methods). Data are presented as
mean+s.e.mean of eight rats.
pharmacokinetics, CSF and bECF neuropharmacokinetics.
The rat models used have numerous advantages: (1) as each rat
serves as its own control, interindividual variability (as occurs
when composite values of individual rats killed at di€erent
time points after drug administration are used) is minimized
and thus the number of experimental animals needed to obtain
detailed kinetic data is reduced; (2) long sampling protocols
can be used (up to 30 h in the present study) in determining
drug kinetics; and (3) microdialysis permits monitoring of
pharmacodynamically relevant bECF compartment and
simultaneous sampling from more than one brain region.
Microdialysis techniques result in minimal pertubation to the
system measured; CSF sampling, however, does involve the
withdrawal of ¯uid. The volume of ¯uid collected represents
less than 12% of the total CSF volume, and since the rate of
secretion is 2 ml min71, then the volume would be replaced in
10 min (Baudrie et al., 1990; Davson & Segal, 1995). The
maximum rate of sampling of 20 ml every 20 min is
The CSF to serum concentration ratio rises to a steady-state
value (0.62) that is greater than the unbound to total serum
concentration ratio. This can be explained by the fact that the
hypothesis that the free serum drug concentration is the tissue
exchangeable drug concentration is incorrect; indeed,
a
number of studies have demonstrated that serum protein
bound drug is often available for transport into the brain
(Urien et al., 1987; Pardridge et al., 1983; Cornford et al., 1992;
Lolin et al., 1994). The CSF to serum concentration ratio at
steady state is substantially less than the total brain to serum
concentration ratio at steady state, which in rats is
approximately two (Walton et al., 1996). This suggests that
there is signi®cant binding and/or intracellular accumulation in
the brain that could act as a potential reservoir.
Due to the peculiar serum pharmacokinetics of lamotrigine
± the serum concentration decays only slowly after approxi-
British Journal of Pharmacology, vol 130 (2)

M.C. Walker et al
Central and peripheral kinetics of lamotrigine in rat
247
mately 8 h ± it was possible for us to calculate the in vivo
recovery of the microdialysis probes using the no-net-¯ux
method (Lonnroth et al., 1987; de-Lange et al., 1997). This is
dependent on an unvarying bECF concentration of the drug
during the study ± a fact that we con®rmed by comparing the
dialysate concentration before and after the no-net-¯ux study.
The in vitro microdialysis probe recovery was approximately
double the calculated in vivo microdialysis probe recovery.
There are a number of factors that may contribute to this (de-
Lange et al., 1997). The in vitro recovery is performed in a
solution of the substrate, which is very di€erent from the in
vivo situation in which the microdialysis probe is surrounded
by brain tissue, which may inhibit free di€usion. Also the
microdialysis probes are in place for 2 days before the
experiment takes place; this is to allow recovery of the animals
from surgery (Patsalos et al., 1992). During this time, however,
proteins bind to the dialysis membrane and are likely to
decrease substantially the permeability of the membrane; in
vivo microdialysis probe recoveries after a similar period of
time after implantation have been shown to be up to 2 fold
lower than the in vitro recoveries (de-Lange et al., 1994). The
marked di€erence between in vivo and in vitro recoveries
emphasizes the importance of calculating in vivo recoveries in
order to get accurate measures of the bECF concentration.
Correcting for estimated in vivo recovery, resulted in bECF
concentration measurements that at steady-state approach the
free serum concentration. These values are signi®cantly smaller
than the measured CSF concentrations at steady-state,
emphasizing the di€erence between these two compartments;
thus lamotrigine concentrations within the CSF compartment
are not always an accurate indicator of concentrations in the
bECF compartment. Furthermore the higher concentrations in
the CSF suggest that the choroid plexus is playing a permissive
role, perhaps actively transporting the drug.
The concentrations in hippocampus and frontal cortex were
identical; this contrasts with studies in which we found higher
concentrations of phenytoin in the hippocampus than in the
frontal cortex (Walker et al., 1996), and higher concentrations
of vigabatrin in the frontal cortex than in the hippocampus
(Patsalos et al., 1999). The brain distribution of AEDs may
partly explain spectrum of antiepileptic activity, and it is
interesting that lamotrigine and phenytoin, two drugs that
have similar mechanisms of action, should have di€erent
spectrums of antiepileptic activity ± lamotrigine, but not
phenytoin is e€ective in absences.
This study emphasizes the importance of determining both
the peripheral (serum) and central (CSF and bECF) kinetics of
a drug using techniques that permit concurrent sampling over
an extensive period. Such data are essential if appropriate
dosing strategies are to be employed to study the pharmaco-
dynamics of a drug and also to aid appropriate interpretation
of experimental data.
We wish to thank the Wellcome Trust for supporting the work of
M.C. Walker and H. Perry, Mr A.A. Elyas for his invaluable
assistance in the h.p.l.c. analysis and Glaxo-Wellcome for the supply
of lamotrigine.
References
BAUDRIE, V., ROULLET, J.B., GOUREAU, Y., CHAOULOFF, F. &
ELGHOZI, J.L. (1990). Determination of cerebrospinal ¯uid
production rate using a push-pull perfusion procedure in the
conscious rat. Fundam. Clin. Pharmacol., 4, 269 ± 274.
CORNFORD, E.M., YOUNG, D., PAXTON, J.W. & SOFIA, R.D. (1992).
Blood-brain barrier penetration of felbamate. Epilepsia, 33,
944 ± 954.
CRUMRINE, R.C., BERGSTRAND, K., COOPER, A.T., FAISON, W.L. &
COOPER, B.R. (1997). Lamotrigine protects hippocampal CA1
neurons from ischemic damage after cardiac arrest. Stroke, 28,
2230 ± 2236.
HUTSON, P.H., SARNA, G.S., KANTAMANENI, B.D. & CURZON, G.
(1985). Monitoring the e€ect of a tryptophan load on brain
indole metabolism in freely moving rats by simultaneous
cerebrospinal ¯uid monitoring and brain dialysis. J. Neurochem.,
44, 1266 ± 1274.
LOLIN, Y.I., RATNARAJ, N., HJELM, M. & PATSALOS, P.N. (1994).
Antiepileptic drug pharmacokinetics and neuropharmacoki-
netics in individual rats by repetitive withdrawal of blood and
cerebrospinal ¯uid: phenytoin. Epilepsy Res., 19, 99 ± 110.
LONNROTH, P., JANSSON, P.A. & SMITH, U. (1987). A microdialysis
method allowing characterization of intercellular water space in
humans. Am. J. Physiol., 253, E228 ± E231.
DAVSON, H. & SEGAL, M.B. (1995). Physiology of the CSF and blood
brain barriers. Florida: CRC Press.
LOSCHER, W.
& FREY, H.H. (1984). Kinetics of penetration of
DE-LANGE, E.C., BOUW, M.R., MANDEMA, J.W., DANHOF, M., DE-
BOER, A.G. & BREIMER, D.D. (1995). Application of intracer-
ebral microdialysis to study regional distribution kinetics of
drugs in rat brain. Br. J. Pharmacol., 116, 2538 ± 2544.
DE-LANGE, E.C., DANHOF, M., DE-BOER, A.G. & BREIMER, D.D.
common antiepileptic drugs into cerebrospinal ¯uid. Epilepsia,
25, 346 ± 352.
MAYER, S., MAICKEL, R.P. & BRODIE, B.B. (1959). Kinetics of
penetration of drugs and other foreign compounds into the
cerebrospinal ¯uid and brain. J. Pharmacol. Exp. Ther., 127,
205 ± 211.
(1994). Critical factors of intracerebral microdialysis as
a
technique to determine the pharmacokinetics of drugs in rat
brain. Brain Res., 666, 1 ± 8.
DE-LANGE, E.C., DANHOF, M., DE-BOER, A.G. & BREIMER, D.D.
(1997). Methodological considerations of intracerebral micro-
dialysis in pharmacokinetic studies on drug transport across the
blood-brain barrier. Brain Res. Rev., 25, 27 ± 49.
MCGEER, E.G. & ZHU, S.G. (1990). Lamotrigine protects against
kainate but not ibotenate lesions in rat striatum. Neurosci. Lett.,
112, 348 ± 351.
MILLER, A.A., WHEATLEY, P., SAWYER, D.A., BAXTER, M.G. &
ROTH, B. (1986). Pharmacological studies on lamotrigine, a novel
potential antiepileptic drug: I. Anticonvulsant pro®le in mice and
rats. Epilepsia, 27, 483 ± 489.
DICKINS, M., SAWYER, D.A., MORLEY, T.J.
& PARSONS, D.N.
(1995). Lamotrigine: Chemistry and biotransformation. In
Antiepileptic drugs. ed. Levy, R.H., Mattson, R.H. & Meldrum,
B.S. pp. 871 ± 875. New York: Raven Press.
NAGAKI, S., RATNARAJ, N. & PATSALOS, P.N. (1999). Blood and
cerebrospinal ¯uid pharmacokinetics of primidone and its
primary pharmacologically active metabolites, phenobarbital
and phenylethylmalonamide in the rat. Eur. J. Drug Metab.
Ph., 24, 255 ± 264.
PARDRIDGE, W.M., SAKIYAMA, R. & FIERER, G. (1983). Transport
of propranolol and lidocaine through the rat blood-brain barrier.
Primary role of globulin-bound drug. J. Clin. Invest., 71, 900 ±
908.
DOHENY, H.C., RATNARAJ, N., WHITTINGTON, M.A., JEFFERYS,
J.G. & PATSALOS, P.N. (1999). Blood and cerebrospinal ¯uid
pharmacokinetics of the novel anticonvulsant levetiracetam (ucb
L059) in the rat. Epilepsy Res., 34, 161 ± 168.
FITTON, A. & GOA, K.L. (1995). Lamotrigine. An update of its
pharmacology and therapeutic use in epilepsy. Drugs, 50, 691 ±
713.
British Journal of Pharmacology, vol 130 (2)

248
M.C. Walker et al
Central and peripheral kinetics of lamotrigine in rat
PARSONS, D.N., DICKINS, M. & MORLEY, T.J. (1995). Lamotrigine:
absorption, distribution and excretion. In Antiepileptic drugs. ed.
Levy, R.H., Mattson, R.H. & Meldrum, B.S. pp. 877 ± 881. New
York: Raven Press.
THOMAS, S.A. & SEGAL, M.B. (1998). The transport of the anti-HIV
drug, 2',3'-didehydro-3'-deoxythymidine (D4T), across the
blood-brain and blood-cerebrospinal ¯uid barriers. Br. J.
Pharmacol., 125, 49 ± 54.
PATSALOS, P.N., ABED, W.T., ALAVIJEH, M.S. & O'CONNELL, M.T.
(1995). The use of microdialysis for the study of drug kinetics:
some methodological considerations illustrated with antipyrine
in rat frontal cortex. Br. J. Pharmacol., 115, 503 ± 509.
PATSALOS, P.N., ALAVIJEH, M.S., SEMBA, J. & LOLIN, Y.I. (1992). A
freely moving and behaving rat model for the chronic and
simultaneous study of drug pharmacokinetics (blood) and
neuropharmacokinetics (cerebrospinal ¯uid): hematological and
biochemical characterization and kinetic evaluation using
carbamazepine. J. Pharmacol. Toxicol. Methods, 28, 21 ± 28.
PATSALOS, P.N., TONG, X. & RATNARAJ, N. (1999). Intracerebral
microdialysis shows that vigabatrin does not exhibit ubiquitous
brain extracellular distribution. Seizure, 8, 372.
URIEN, S., PINQUIER, J.L., PAQUETTE, B., CHAUMET, R.P.,
KIECHEL, J.R. & TILLEMENT, J.P. (1987). E€ect of the binding
of isradipine and darodipine to di€erent plasma proteins on their
transfer through the rat blood-brain barrier. Drug binding to
lipoproteins does not limit the transfer of drug. J. Pharmacol.
Exp. Ther., 242, 349 ± 353.
WALKER, M.C., ALAVIJEH, M.S., SHORVON, S.D. & PATSALOS, P.N.
(1996). Microdialysis study of the neuropharmacokinetics of
phenytoin in rat hippocampus and frontal cortex. Epilepsia, 37,
421 ± 427.
WALKER, M.C. & SANDER, J.W. (1999). Lamotrigine. In Antiepileptic
Drugs: pharmacology and therapeutics. ed. M.J. Eadie, M.E. &
Vajda, F. pp. 331 ± 338. Berlin: Springer-Verlag.
PAXINOS, G. & WATSON, C. (1986). The rat brain in stereotaxic co-
ordinates. 2nd edition. San Diego: Academic Press.
RATAUD, J., DEBARNOT, F., MARY, V., PRATT, J. & STUTZMANN,
J.M. (1994). Comparative study of voltage-sensitive sodium
channel blockers in focal ischaemia and electric convulsions in
rodents. Neurosci. Lett., 172, 19 ± 23.
WALKER, M.C., TONG, X., BROWN, S., SHORVON, S.D. & PATSALOS,
P.N. (1998). Comparison of single- and repeated-dose pharma-
cokinetics of diazepam. Epilepsia, 39, 283 ± 289.
WALTON, N.Y., JAING, Q., HYUN, B. & TREIMAN, D.M. (1996).
Lamotrigine vs phenytoin for treatment of status epilepticus:
comparison in an experimental model. Epilepsy Res., 24, 19 ± 28.
WIARD, R.P., DICKERSON, M.C., BEEK, O., NORTON, R. & COOPER,
B.R. (1995). Neuroprotective properties of the novel antiepileptic
lamotrigine in a gerbil model of global cerebral ischemia. Stroke,
26, 466 ± 472.
SECHI, G.P., PETRUZZI, V., ROSATI, G., TANCA, S., MONACO, F.,
FORMATO, M., RUBATTU, L.
& DE-RIU, P. (1989). Brain
interstitial ¯uid and intracellular distribution of phenytoin.
Epilepsia, 30, 235 ± 239.
SEMBA, J., CURZON, G. & PATSALOS, P.N. (1993). Antiepileptic drug
pharmacokinetics and neuropharmacokinetics in individual rats
by repetitive withdrawal of blood and cerebrospinal ¯uid:
milacemide. Br. J. Pharmacol., 108, 1117 ± 1124.
SMITH, S.E. & MELDRUM, B.S. (1995). Cerebroprotective e€ect of
lamotrigine after focal ischemia in rats. Stroke, 26, 117 ± 121.
WILDER, B.J., RAMSAY, R.E., WILLMORE, L.J., FEUSSNER, G.F.,
& SHUMATE JR. J.B. (1977). Ecacy of
PERCHALSKI, R.J.
intravenous phenytoin in the treatment of status epilepticus,
kinetics of central nervous system penetration. Ann. Neurol., 1,
511 ± 518.
(Recieved November 24, 1999
Revised February 21, 2000
Accepted March 1, 2000)
British Journal of Pharmacology, vol 130 (2)