APPENDIX _- SUPPLEMENTAL MATERIAL
Bioanalytical method for the measurement of gabapentin concentrations in plasma:
Blood samples (100μl) were taken at the pre-defined time points up to 5 hours post-dose,
namely 0, 2, 5, 15, 30, 60, 120, 180, 240 and 300 min. Plasma samples (50μl) were
obtained by centrifugation at 4°C for 10 min and stored at -80°C until analysis.
Gabapentin concentration in plasma was subsequently analysed by HPLC using precolumn derivatisation. Gabapentin and the internal standard, i.e., 1-(aminomethyl)
cycloheptaneacetic acid, were allowed to react with 2, 4, 6-trinitrobenzenesulfone acid to
form trinitrophenyl derivatives, which were then extracted with toluene, evaporated to
dryness and reconstituted before injection. Analytes were resolved on a C 18 reverse phase
column using isocratic conditions. Mobile phase consisted of 58% acetonitrile in water
containing 0.5% acetic acid. Ultraviolet absorbance was monitored at 35min.
Quantification of gabapentin levels was based on the peak-height ratio. The lower limit of
detection for gabapentin was 0.02 μg/ml (52, 53)
Simulation of gabapentin concentrations: The concentration vs. time course of
gabapentin in plasma was obtained by simulating the drug levels at the time points when
pharmacodynamic measurements were recorded. A two-compartment pharmacokinetic
model with 1st-order absorption and elimination processes was used for the simulations
(16). This model was based on data from two different experiments in Sprague-Dawley
rats. In the first experiment, gabapentin was administered orally to conscious rats at doses
of 0, 10, 100, 300 mg/kg in a formalin-induced hypersensitivity experiment similar to the
1
protocol described in the current manuscript, which in turn was also based on a
previously published experimental protocol (2, 6). Each treatment group consisted of
three rats per dose level, with each animal contributing with four samples over a period of
up to 6 h post-dose. The second experiment was a microdialysis protocol, in which rats
received intravenous doses of 50 mg/kg gabapentin (n=63). Each animal contributed with
eight samples over a period of up to 24 h post-dose (51).
Analytical solution for the two-compartment pharmacokinetic model used in the
simulations: As described by main text, gabapentin plasma concentrations were derived
from equation 1:
C
k a FD  (k 21 1 )e 1t
(k 21 k a )e  Ka t 
(k 21 2 )e 2t


(


V1  (k a  1 )( 2  1 ) (k a  2 )( 1  2 ) (1  k a )( 2  k a ) 
where ka represents the absorption rate constant, F = bioavailability, D = dose, V1 =
volume of distribution of the central compartment, k12 = micro-rate constant describing
the transfer of gabapentin from the central to peripheral compartment, k21 = micro-rate
constant describing the transfer of gabapentin from the peripheral to central compartment.
The two macro-constants λ1 and λ2 (corresponding to the initial and terminal slopes
representing bi-exponential decline, respectively) may be further derived as follows:
λ1=
A  A2  4k 21k el
2
2
λ2=
A  A2  4k 21k el
2
where A = k12+k21+kel and kel = elimination rate constant (from the plasma compartment).
The relationship between the primary model parameters and micro-rate constants are
summarised below:
k el 
Cl
Q
Q
, k12  , k 21 
V1
V1
V2
where kel = elimination rate constant, k12 = micro-rate constant describing the transfer of
gabapentin from the central to peripheral compartment, k21 = micro-rate
constant
describing the transfer of gabapentin from the peripheral to central compartment, V1 =
central volume of distribution, V2 = peripheral volume of distribution, Cl = clearance
from plasma, Q = intercompartmental clearance.
3
Figure S1: Assessment of correlations between successive measurements at consecutive
time interval. The pain frequency at each time interval is plotted versus the frequency at
the preceding measurement. The frequency of counts at any particular time interval is
highly correlated with the values observed in the preceding interval indicating structural
correlation between successive measurements.
4
Table 1S: Comparison of Experimental findings for gabapentin in various published preclinical and clinical studies.
No.
Author
Experimental
Model
1.
Shannon et al.,
2005
Formalininduced pain in
rats
2.
Iyengar et al.,
2004
Formalininduced pain in
rats
3.
Hama and
Sagen, 2007
Rat model of
acute
neuropathic pain
by spinal
compression
injury
Study Protocol
Main findings
Comparison with pre-clinical experiments
Comparison of anticonvulsants with In rats efficacy was seen across
different mechanisms of action in the the dose range from 30-300
formalin test of persistent pain in rats mg/kg.
and mice. Gabapentin doses of 30 The minimal effective dose
300 mg/kg were tested after i.p.
(MED) in rats was 30 mg/kg for
administration.
the second peak of pain, whilst
for locomotor activity the MED
in mice was 100 mg/kg
Comparison of the effects of
analgesic agents, such as uptake
inhibitors, antidepressants and
anticonvulsants on the attenuation of
formalin-induced late phase pawlicking behaviour. Gabapentin doses
of 10, 30 and 50 mg/kg were tested
after i.p. administration.
A placebo controlled 12-week study.
A number of compounds among
which opioid analgesics,
antidepressants, anticonvulsants were
tested. Gabapentin doses of
10/30/100mg/kg were tested after i.p.
administration.
Comparison
with current
work and other
remarks
Drug effects on
the second peak
ranged between
30-100 mg/kg
Gabapentin attenuated paw
licking behaviour at the tested
doses. Maximum effect was
reached at 50 mg/kg.
In our
experiments,
analgesia was
observed at 100
mg/kg.
Gabapentin dose-dependently
reversed mechanical
hypersensitivity. The AD50
(antinoceptive dose) was 26 (1642) mg/kg. Maximum effect was
observed 90 min after injection.
IC50 estimates
were higher than
the exposure
range observed
after
administration of
doses between 10
5
No.
Author
Experimental
Model
Study Protocol
Main findings
4.
Whiteside et
al., 2004
Spinal nerve
ligation rat
model, clinical
data
Comparison of human Cmax at MED
at daily maintenance dose (1800 mg)
to rat MED based on published
literature.
5.
Yoon and
Yaksh, 1999
Formalininduced pain in
rats
6.
Hurley et al.,
2002
Rat model of
carrageenaninduced
inflammation
7.
Whiteside et
al., 2004
Rat model of
incision pain
The anti-hyperalgesic effects of
gabapentin were tested at doses of
10, 30, 100 and 300 mg/kg after i.p.
administration of gabapentin alone
and in combination with ibuprofen.
An isobolographic analysis was used
to analyse drug-drug interaction.
Gabapentin, pregabalin or naproxen
were administered alone or in
combination as oral gavage to rats.
An isobolo-graphic analysis was
used to characterise the interaction.
Gabapentin was administered at
doses ranging from 3-300mg/kg.
A number of analgesic drugs such as
indomethacin, gabapentin and
morphine were compared.
Gabapentin doses were 10, 30
and100 mg/kg.
The concentration corresponding
to the MED (100 mg/kg) in rats
was 191.5 nM, whilst plasma
levels in humans reach 69.7 nM
after maintenance doses of 26
mg/kg
The ED50 values for gabapentin
and ibuprofen were, respectively,
88 (51-141) mg/kg and 19 (7–50)
mg/kg.
The ED50 values for gabapentin,
pregabalin and naproxen were
respectively 19.2 (5.5-43.1)
mg/kg, 6 (2.3–10.0) mg/kg and
0.48 (0.05-1.3) mg/kg.
The MED for mechanical
hyperalgesia was 30 mg/kg, with
ED50 estimates of 11.3 mg/kg.
For tactile allodynia the MED
was 11 mg/kg, with ED50 of
3.4mg/kg.
Comparison
with current
work and other
remarks
and100 mg/kg.
IC50 was
estimated to be
43 nM with a
coefficient of
variation of 40%.
The IC50 values
corresponded to
exposures higher
than 100mg/kg.
In this model
parameter
estimates suggest
higher potency
for gabapentin as
compared to the
current analysis.
In this model
parameter
estimates suggest
higher potency
for gabapentin as
compared to the
6
No.
Author
Experimental
Model
8.
Todorovic et
al., 2003
The radiant heat
rat model of
neuropathic
pain.
10.
Whiteside et
al., 2004
11.
Lockwood et
al., 2003
Study Protocol
Main findings
Anticonvulsants were injected
The ED50 estimate was 80
intradermally into the peripheral
μg/100 ml or 4.67 nM
receptive fields of sensory neurons in
the hind paws of adult rats, and paw
withdrawal latency measured.
Gabapentin (5-170μg), phenytoin
(0.1-3 μg), carbamazepine (0.1-2 μg)
and ethosuximide (140-1400 μg)
were evaluated. Dose–response data
were fit to the Hill equation.
Comparison with clinical and translational experiments
Spinal nerve
Comparison of human Cmax
The rat MED was 191.5 nM as
ligation rat
following daily maintenance doses of compared to 69.7 nM in humans
model, clinical
1800mg to rat MED based on
data
published literature.
A phase 3 study
in patients with
neuropathic pain
This was a placebo-controlled,
double blinded study where patients
were randomised to placebo or
gabapentin. A PKPD model was
fitted to data and further used to
simulate the MED for an
investigational compound
(pregabalin) based on in vitro
potency information.
The IC50 estimated as 31.2 nM.
Comparison
with current
work and other
remarks
current analysis.
The 10-fold
difference in
potency may be
explained by the
different routes
of administration.
IC50 was
estimated to be
43 nM with a
coefficient of
variation of 40%.
The estimated
IC50 value is in
the same order of
magnitude of the
values observed
in the current
analysis.
7
8