Fortuna 2011

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 9

Research article

Received: 26 March 2011, Accepted: 8 June 2011 Published online in Wiley Online Library: 21 July 2011

(wileyonlinelibrary.com) DOI 10.1002/bmc.1670

A chiral liquid chromatography method for the


simultaneous determination of oxcarbazepine,
eslicarbazepine, R-licarbazepine and other new
chemical derivatives BIA 2–024, BIA 2–059 and
BIA 2–265, in mouse plasma and brain
Ana Fortunaa,b, Gilberto Alvesb,c, Anabela Almeidaa,b, Bruno Lopesa,
Amílcar Falcãoa,b* and Patrício Soares-da-Silvad
ABSTRACT: Recently, in silico models have been developed to predict drug pharmacokinetics. However, before application,
they must be validated and, for that, information about structurally similar reference compounds is required. A chiral liquid
chromatography method with ultraviolet detection (LC-UV) was developed and validated for the simultaneous quantification
of BIA 2–024, BIA 2–059, BIA 2–265, oxcarbazepine, eslicarbazepine (S-licarbazepine) and R-licarbazepine in mouse plasma and
brain. Compounds were extracted by a selective solid-phase extraction procedure and their chromatographic separation was
achieved on a LiChroCART 250–4 ChiraDex column using a mobile phase of water–methanol (92:8, v/v) pumped at 0.7mL/min.
The UV detector was set at 235nm. Calibration curves were linear (r 2 ≥0.996) over the concentration ranges of 0.2–30mg/mL for
oxcarbazepine, eslicarbazepine and R-licarbazepine; 0.2–60mg/mL for the remaining compounds in plasma; and 0.06–15mg/mL
for all the analytes in brain homogenate. Taking into account all analytes at these concentration ranges in both matrices, the
overall precision did not exceed 9.09%, and the accuracy was within 14.3%. This LC-UV method is suitable for carrying out
pharmacokinetic studies with these compounds in mouse in order to obtain a better picture of their metabolic pathways and
biodistribution. Copyright © 2011 John Wiley & Sons, Ltd.

Keywords: chiral liquid chromatography; bioanalytical validation; mouse plasma and brain; oxcarbazepine; eslicarbazepine

Introduction low-toxicity properties, emphasizing the importance of this


evaluation in the initial phases of DDD (Lipinski, 2000; Yengi
At the initial phases of the drug discovery and development et al., 2007; Tsaioun et al., 2009). In this context, basic research
(DDD) process, new chemical entities (NCEs) have traditionally
been screened over several decades based on their biological
activity and therapeutic effects using studies performed in
whole animal models (Nef, 2001). However, since the early * Correspondence to: Amílcar Falcão, Pharmacology Department, Faculty of
1990s, the challenges facing the pharmaceutical industries have Pharmacy, University of Coimbra Pólo das Ciências da Saúde, Azinhaga
been tremendous at every step of DDD because of the exponen- de Santa Comba 3000-548 Coimbra, Portugal. E-mail:acfalcao@ff.uc.pt
tial increase in the number of compounds emerging from com- a
Pharmacology Department, Faculty of Pharmacy, University of Coimbra,
binatorial chemistry and high-throughput screening methods, 3000-548 Coimbra, Portugal
which has increased the demand for screening of NCEs using a
b
wide range of biological assays. A variety of in silico and Centre for Neurosciences and Cell Biology, University of Coimbra, 3004-517
Coimbra, Portugal
in vitro approaches have been developed and combined to iden-
tify the most potent and selective drug candidate, reducing the c
Health Sciences Research Centre (CICS), University of Beira Interior, 6200-506
costs and time required for new drugs to reach the market. Nev- Covilhã, Portugal
ertheless, this strategy of screening thousands of compounds d
solely for potency and selectivity has not been as efficient as Department of Research and Development, BIAL, 4745-457 S. Mamede do
Coronado, Portugal
expected, because a large number of NCEs fail during clinical
trials and do not reach the market. Problems with absorption, Abbreviations used: ADME, absorption, distribution, metabolism and ex-
distribution, metabolism and excretion (ADME) or toxicity have cretion; BIA 2–024, 10,11-dihydro-10-hydroxyimino-5H-dibenz[b,f]azepine-
been identified as the major reasons for failure of drug 5-carboxamide; BIA 2–059, (+)-(R)-10-acetoxy-10,11-dro-5H-dibenz[b,f]
azepine-5-carboxamid; BIA 2–265, 10-(hydroxymethyl)-10-nitro-10,11-
candidates in clinical development. About 50% of promising dihydro-5H-dibenzo[b,f]azepine-5-carboxamide; DDD, drug discovery
drug candidates do not progress through the DDD process and development; NCE, new chemical entity; OXC, oxcarbazepine; R-Lic,
because they do not possess the required pharmacokinetic or R-licarbazepine; S-Lic, S-licarbazepine.
384

Biomed. Chromatogr. 2012; 26: 384–392 Copyright © 2011 John Wiley & Sons, Ltd.
Determination of eslicarbazepine and derivatives in mouse

using in silico and in vitro assays with high throughput and low
costs must be extensively performed. These tools decrease the
turnaround time for delivery of information to medicinal che-
mists during the lead optimization phase. Such methodologies
should also be combined in an iterative way in order to prioritize
compounds with acceptable therapeutic potential and pharma-
cokinetic properties for advance to more expensive in vivo
testing phases (Pauli et al., 2008). Effectively, despite the consid-
erable changes recently introduced in the DDD process, in vivo
experimental studies using laboratory animal models remain de-
cisive in the development of new drugs, but they are now usu-
ally required only at later phases of pre-clinical development.
Pre-clinical studies in whole animal models are considered to
be essential for better assessing the pharmacology, toxicity
and pharmacokinetic properties of new drug candidates, with
the aim of extrapolating the results to humans. In fact, only com-
pounds demonstrating acceptable pharmacokinetic and safety
profiles in animals should progress to the first clinical trials for
in-depth studies (Bohets et al., 2001; Caldwell et al., 2001).
Accordingly, there is increasing interest in optimizing computer
and statistical models that simulate pharmacokinetics and toxicity
in the body (Schmitt and Willmann, 2004; Parrot et al., 2005;
Carlson and Fisher, 2008; Sui et al., 2008). However, for construc-
tion of such in silico pharmacostatistical models, adequate infor-
mation is required. In addition, before such tools can be routinely
used, they have to be validated by comparing the predicted
data with the information obtained experimentally for known Figure 1. Chemical structures of (+)-(R)-10-acetoxy-10,11-dihydro-
compounds. 5H-dibenz[b,f]azepine-5-carboxamide (BIA 2-059), 10,11-dihydro-
Over the last 40years several dibenz[b,f]azepine-5-carboxa- 10-hydroxyimino-5Hdibenz[b,f]azepine-5-carboxamide (BIA 2-024),
mide derivatives have been empirically synthesized, and three 10-(hydroxymethyl)-10-nitro-10,11-dihydro-5H-dibenzo[b,f]azepine-
generations of structurally related but distinct antiepileptic 5-carboxamide (BIA 2-265), oxcarbazepine (OXC), eslicarbazepine
(S-licarbazepine; S-Lic), R-licarbazepine (R-Lic), and carbamazepine-10,
drugs are marketed: carbamazepine, oxcarbazepine (OXC) and
11-epoxide used as internal standard (IS).
eslicarbazepine acetate (Neels et al., 2004; Luszczki, 2009). How-
ever, many other dibenz[b,f]azepine-5-carboxamide derivatives
have been demonstrated to be devoid of antiepileptic potential chromatography elution conditions used, whenever possible,
after their evaluation against the gold standard seizure animal during the DDD in pharmacokinetic departments.
models. Thus, our research group aims to better understand the
reasons for the success or failure of many dibenz[b,f]azepine-
5-carboxamide compounds, taking advantage of in silico ap- Experimental
proaches specifically developed for these compounds. For this
purpose we aimed to obtain more pharmacokinetic data about Chemicals and reagents
some dibenz[b,f]azepine-5-carboxamide derivatives that have OXC (lot number FO40003), S-Lic (lot number 2070-2-1), R-Lic (lot num-
not been extensively evaluated: BIA 2–024 {10,11-dihydro-10- ber PC040414), BIA 2–059 (lot number PC040831), BIA 2–024 (lot number
hydroxyimino-5H-dibenz[b,f]azepine-5-carboxamide}, BIA 2–059 DL980709), BIA 2–265 (lot number PC050704) and carbamazepine-10,11-
{ ( + ) - ( R)-10-acetoxy-10,11-dihydro-5H-dibenz[b,f]azepine-5 - car- epoxide (lot number PC030916), used as internal standard (IS), were
boxamide}, and BIA 2–265 {10-(hydroxymethyl)-10-nitro-10,11- kindly supplied by BIAL (Porto, Portugal; Fig. 1). Methanol (LC gradient
dihydro-5H-dibenzo[b,f]azepine-5-carboxamide} (Fig. 1). However, grade) was purchased from Fisher Scientific (Leicestershire, UK) and
to date, there have been no bioanalytical methods developed to ultra-pure water (LC grade, >15MΩ) was obtained by means of a Milli-Q
simultaneously quantify these compounds. Therefore, the avail- water apparatus from Millipore (Milford, MA, USA). Ethyl acetate, sodium
dihydrogen phosphate dihydrate, di-sodium hydrogen phosphate dihy-
ability of a reliable bioanalytical method to quantify BIA 2–024,
drate and hydrochloric acid fuming 37% were purchased from Merck KGaA
BIA 2–059 and BIA 2–265 would be required. In addition, to
(Darmstadt, Germany).
get a better picture of the metabolic pathways of these com-
pounds and to assess their probable metabolism to well-known
compounds such as OXC, eslicarbazepine (S-licarbazepine; S-Lic), Chromatographic instrumentation and analytical conditions
and R-licarbazepine (R-Lic), we decided also to consider them as
The chromatographic analysis was carried out using a BAS-480 liquid
analytes (Fig. 1). Obviously, this fact prompted us to develop and
chromatograph equipped with a PM-80 pump, a Rheodyne manual
fully validate the first liquid chromatography (LC) method able
injector with a 20mL loop, a BAS UV-116 UV–vis detector, a BAS LC-
to quantify simultaneously BIA 2–024, BIA 2–059, BIA 2–265, 22 C temperature controller, a BAS DA-5 chromatography control
OXC, S-Lic and R-Lic, in mouse plasma and brain. Moreover, and a data system interface (all from Bioanalytical Systems, West
we chose to develop the new LC method employing ultra- Lafayette, IN, USA). Data collection and integration were achieved by
violet (UV) detection and isocratic conditions which are, means of BAS Chromgraph Control and Chromgraph Report software
respectively, the most usual detection system and the simplest version 2.30. Chromatographic separation of OXC, S-Lic, R-Lic, BIA
385

Biomed. Chromatogr. 2012; 26: 384–392 Copyright © 2011 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/bmc
A. Fortuna et al.

2–059, BIA 2–024, BIA 2–265 and the IS was performed at 35 C on accommodation and care of laboratory animals and the experimental
a LiChroCART 250–4 ChiraDex (b-cyclodextrin, 5mm) column protected procedures were approved by the Portuguese Veterinary General
by a LiChroCART 4–4 ChiraDex (b-cyclodextrin, 5mm) guard column Division.
purchased from Merck KGaA (Darmstadt, Germany) and using a mobile
phase composed of water–methanol (92:8, v/v), which is delivered at a
flow-rate of 0.7mL/min. The mobile phase was filtered through a 0.45mm Extraction procedure
filter and ultrasonically degassed prior to use. The wavelength of
detection was set at 235nm. Samples were prepared by solid-phase extraction (SPE) process using the
OasisW HLB (30mg, 1mL) cartridges from Waters, Milford, MA, USA. At
the day of analysis, after plasma had been thawed, 300mL were added
Preparation of stock solutions, calibration samples and to 700mL of 0.1 M sodium phosphate buffer (pH 5) and spiked with 10mL
quality control samples of the IS working solution. Samples were mixed and loaded into the
SPE cartridges, previously conditioned with 1mL of methanol, 1mL of
Stock solutions of OXC, S-Lic, R-Lic, BIA 2–059, BIA 2–024 and BIA 2–265 acetonitrile and 1mL of water–acetonitrile (95:5, v/v). After sample elu-
were individually prepared in acetonitrile at concentrations of 10mg/mL tion, the loaded cartridges were submitted to 30 kPa and washed four
and appropriately diluted to obtain solutions of 1mg/mL and 100mg/mL. times with 1mL of water–methanol (90:10, v/v), then the sorbent was
Considering the IS, its stock solution of 1mg/mL was prepared by dissol- dried under air flow for 5min. Thereafter, analytes were eluted with 1
ving the appropriate amount in acetonitrile and the working solution of mL of ethyl acetate under a gentle vacuum and the eluates were evapo-
200mg/mL was obtained daily by appropriately diluting the stock solu- rated to dryness under a nitrogen stream at 45 C. The residues were
tion with acetonitrile. Drug working solutions were combined spiking reconstituted in 100mL of LC mobile phase, vortexed for approximately
solutions obtained by mixing stock and dilution solutions with acetoni- 1min and placed in an ultrasonic bath at room temperature for approx-
trile. The combined spiking solutions used for the preparation of calibra- imately 1min. Following this, the reconstituted extracts were trans-
tion standards in plasma presented final concentrations of 6, 12, 36, 90, ferred to a 0.22mm Spin-X centrifugal filter, centrifuged at 13400rpm
300 and 900mg/mL for OXC, S-Lic and R-Lic, and 6, 12, 36, 150, 600 and for 2min and 20 mL of the final filtered samples were injected into the
1800mg/mL for BIA 2–059, BIA 2–024 and BIA 2–265. On the other hand, LC system.
the corresponding combined spiking solutions used for the preparation
After thawing, 1mL of the brain homogenate supernatant was directly
of calibration standards in brain homogenate supernatant were pre-
spiked with 10mL of IS and then submitted to the SPE procedure previ-
pared at 6, 12, 36, 140, 500 and 1500mg/mL for all the analytes. Stock, di-
ously described for plasma samples.
lution and working solutions were all refrigerated at 4 C and protected
from light.
Six calibration standards were obtained spiking 10mL of the appropri-
ate working solution into 300mL of blank plasma or 1mL of brain homog- Assay validation
enate supernatant and thereafter used for construction of the calibration Considering the guidelines of one of the main regulatory agencies (US
curves in each matrix. In plasma, the calibration ranges were 0.2–30mg/ DHHS et al., 2001) and the updated recommendations of the Conference
mL for OXC, S-Lic and R-Lic, and 0.2–60mg/mL for BIA 2–059, BIA 2–024 Report of the Washington Conference on ‘analytical methods validation:
and BIA 2–265, whereas in brain homogenate supernatant, the calibra- bioavailability, bioequivalence and pharmacokinetic studies’ (Shah et al.,
tion range was 0.06–15mg/mL for all the analytes. 2000), this method was validated for selectivity, linearity, intra- and inter-
Quality control (QC) samples at low (QC1), middle (QC2) and high (QC3) day accuracy, intra and inter-day precision, lower limit of quantification
concentrations of the calibration ranges were also prepared indepen- (LOQ), recovery and stability.
dently in the studied biological matrices. Thus, plasma (300mL) was Selectivity is described as the ability of a method to discriminate the
spiked to attain 0.5mg/mL for all the compounds in QC1; 15mg/mL for analyte from all potentially interfering substances. It was investigated
OXC, S-Lic and R-Lic, and 30mg/mL for BIA 2–059, BIA 2–024 and BIA by comparing the chromatograms obtained from blank plasma and
2–265 in QC2; and 27mg/mL for OXC, S-Lic and R-Lic, and 54mg/mL for brain homogenate supernatant samples of six different mice with those
BIA 2–059, BIA 2–024 and BIA 2–265 in QC3. Similarly, brain homogenate obtained from the correspondent spiked samples in order to analyze the
supernatant (1mL) was spiked to attain all the compounds at 0.15, 7.5 existence of peaks of endogenous substances interfering with those of
and 13.5mg/mL in QC1, QC2 and QC3, respectively. QC samples were then the analytes.
interpolated from the calibration curves to obtain the concentrations of To evaluate the linearity of the analytical method, calibration curves
each compound. were prepared using six calibration standards and assayed on five sepa-
rate days (n=5). The sensitivity was evaluated by determining the LOQ,
which is defined as the lowest concentration of the calibration curve that
Achievement of mouse plasma and brain homogenates can be measured with acceptable precision and accuracy. It was estab-
In order to obtain drug-free pools of mouse plasma and brain homoge- lished using five extracted samples (n=5), ensuring a percentage of coef-
nate, used as blank matrices during the validation studies, healthy adult ficient of variation (CV%) of not more than 20% and a deviation from
male CD-1 mice with 30–35g and obtained from HarlanTM (Udin, Italy) nominal value in percentage (bias %) within20%.
were sacrificed by cervical dislocation followed by decapitation. Blood The precision of an analytical procedure expresses the closeness of
was immediately collected into heparinized tubes, centrifuged at 4000 agreement between a series of measurements obtained from multiple
rpm for 10min (4 C) and the supernatant plasma stored at 30 C until determinations of the same homogeneous sample, while the accuracy
used. Brain tissues were quickly dissected, weighed, added to 0.1 M so- describes the closeness of experimental results to the nominal value of
dium phosphate buffer pH 5 (4mL to 1g of brain) and, thereafter, homo- the analyte. The intra-day accuracy and precision were determined by
genised using a ThomasW teflon pestle tissue homogenizer. The analyzing five replicates (n=5) of each QC sample on the same day
homogenates were transferred to 10mL glass centrifugation tubes and whereas the inter-day studies were performed on five different days.
centrifuged at 4800rpm for 15min (4 C); the supernatants produced The acceptance criteria for intra- and inter-day precision (expressed as
were stored at 30 C until used. CV%) and accuracy (expressed as percentage bias) were, respectively, a
Mice were kept in local animal facilities under controlled environmen- CV not exceeding 15% and a bias within 15%.
tal conditions (12h light/dark cycle, at 221 C and humidity 505%) The extraction recoveries of all the compounds were determined by
and received a regular standard chow diet (4RF21, Mucedola, Italy) and comparing the peak height ratio (analyte/IS) of extracted samples spiked
tap water ad libitum until used. All the animal experiments were con- with known amounts of the analytes with the peak height ratio of analyt-
ducted in accordance with the European Directive (86/609/EEC) for the ical standards at corresponding concentrations. This procedure was
386

wileyonlinelibrary.com/journal/bmc Copyright © 2011 John Wiley & Sons, Ltd. Biomed. Chromatogr. 2012; 26: 384–392
Determination of eslicarbazepine and derivatives in mouse

repeated five times (n=5) for the three different concentration levels of different temperature and time conditions. The data of QC1 and QC3
QC samples. Recovery of IS was determined in the same way. samples were analyzed before (reference samples) and after being ex-
Experiments were also performed in order to evaluate the stability of posed to the conditions for stability assessment (stability samples) and
the compounds in mobile phase, plasma and brain homogenate under a stability/reference samples ratio of 85–115% was accepted as the sta-
bility criterion (n=5). The short-term (bench-top) temperature stability
was evaluated by keeping the QC samples at room temperature for a pe-
riod of 2h and also at 4 C for 24h. The long-term stability was assessed
at low temperature ( 30 C) for a period of 30days. The post-preparative
stability was measured by reanalyzing extracted QC samples kept at 4 C
for 24h to simulate the time that the sample can be in the auto-sampler.
Finally, the stability was also evaluated after three freeze–thaw cycles.
For that, the aliquots of QC1 and QC3 were stored at 30 C for 24h,
thawed unassisted at room temperature and, when completely thawed,
refrozen for 24h under the same conditions until three cycles had been
completed.

Results and Discussion


Figure 2. Dependence of chromatographic resolution of pairs of
analytes as a function of temperature (BIA 2-059/OXC, OXC/BIA 2-024, LC-UV optimization
and S-Lic/R-Lic). BIA 2-059, (+)-(R)-10-acetoxy-10,11-dihydro-5H-dibenz
[b,f]azepine-5-carboxamide; BIA 2-024, 10,11-dihydro-10-hydroxyimino- A simple and reliable LC method with UV detection was de-
5Hdibenz[b,f]azepine-5-carboxamide; BIA 2-265, 10-(hydroxymethyl)- veloped for the simultaneous quantification of BIA 2–024,
10-nitro-10,11-dihydro-5H-dibenzo[b,f]azepine-5-carboxamide; BIA 2–059, BIA 2–265, OXC, S-Lic and R-Lic, in plasma and brain
OXC, oxcarbazepine; R-Lic, R-licarbazepine; S-Lic, eslicarbazepine of mouse. As we aim to obtain pharmacokinetic information to
(S-licarbazepine). construct and validate in silico models that are useful at early

Figure 3. Typical chromatograms of extracted mouse samples. Blank plasma sample (a), spiked plasma at levels of the limit of quantification (c) and
spiked plasma at levels of the upper limit of the calibration ranges (e); blank supernatant of brain homogenate (b), spiked supernatant of brain homog-
enate at levels of the limit of quantification (d) and spiked supernatant of brain homogenate at levels of the upper limit of the calibration ranges (f). BIA
2-059, (+)-(R)-10-acetoxy-10,11-dihydro-5H-dibenz[b,f]azepine-5-carboxamide; BIA 2-024, 10,11-dihydro-10-hydroxyimino-5Hdibenz[b,f]azepine-5-
carboxamide; BIA 2-265, 10-(hydroxymethyl)-10-nitro-10,11-dihydro-5H-dibenzo[b,f]azepine-5-carboxamide; OXC, oxcarbazepine; R-Lic, R-licarbazepine;
S-Lic, eslicarbazepine (S-licarbazepine); IS, internal standard.
387

Biomed. Chromatogr. 2012; 26: 384–392 Copyright © 2011 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/bmc
A. Fortuna et al.

Table 1. Calibration curve parameters for BIA 2–265, eslicarbazepine (S-licarbazepine, S-Lic), R-licarbazepine (R-Lic), BIA 2–059,
oxcarbazepine (OXC) and BIA 2–024 in mouse plasma and brain homogenate (n=5). The equation of the calibration curve
is y=mx+c, where x is the drug concentration, expressed in mg/mL, and y is the drug to internal standard peak height ratio,
expressed in arbitrary height units. Slope, intercept values and regression coefficients are means

Matrix/drug Calibration standards (µg/mL) Slope (m) Intercept (c) Regression coefficient (r2)
Plasma
BIA 2-265 0.2, 0.4, 1.2, 5, 20, 60 0.1341 0.0030 0.996
S-Lic 0.2, 0.4, 1.2, 3, 10, 30 0.1642 0.0030 0.999
R-Lic 0.2, 0.4, 1.2, 3, 10, 30 0.1266 0.0031 0.997
BIA 2-059 0.2, 0.4, 1.2, 5, 20, 60 0.0887 0.0043 0.999
OXC 0.2, 0.4, 1.2, 3, 10, 30 0.1554 0.0002 0.999
BIA 2-024 0.2, 0.4, 1.2, 5, 20, 60 0.1377 0.0023 0.998
Brain
BIA 2-265 0.06, 0.12, 0.36, 1.4, 5, 15 0.4982 0.0035 0.999
S-Lic 0.06, 0.12, 0.36, 1.4, 5, 15 0.4554 0.0001 0.999
R-Lic 0.06, 0.12, 0.36, 1.4, 5, 15 0.5268 0.0028 0.998
BIA 2-059 0.06, 0.12, 0.36, 1.4, 5, 15 0.3033 0.0002 0.999
OXC 0.06, 0.12, 0.36, 1.4, 5, 15 0.5823 0.0029 0.998
BIA 2-024 0.06, 0.12, 0.36, 1.4, 5, 15 0.4166 0.0007 0.998

saving time and allowing easy switching from one compound


Table 2. Values of precision, expressed by coefficient varia- to another.
tion (CV%), and accuracy, expressed by percentage bias, at As S-Lic and R-Lic are enantiomers, chiral separation has to be
the limit of quantification (LOQ) of BIA 2–265, eslicarbazepine performed either indirectly using derivatization reagents or di-
(S-licarbazepine, S-Lic), R-licarbazepine (R-Lic), BIA 2–059, rectly using chiral selectors as additives in mobile phase or using
oxcarbazepine (OXC) and BIA 2–024 in mouse plasma and chiral stationary phases. The last option was chosen because it is
brain homogenate (n=5) the most convenient and it has been the most frequently used
despite the higher cost and lesser robustness in comparison to
CNominal CExperimentala CV (%) Bias (%)
achiral stationary phases (Alves et al., 2008a). ChiradexW chiral
Matrix/drug (mg/mL) (mg/mL)
stationary phase was chosen and consists of spherical silica par-
Plasma ticles covalently bonded to b-cyclodextrins, which interact differ-
BIA 2-265 0.2 0.2080.013 6.41 3.87 ently with each enantiomer. Indeed S-Lic and R-Lic have been
S-Lic 0.2 0.1890.007 3.51 5.67 separated using this column (Hainzl et al., 2001; Alves et al.,
R-Lic 0.2 0.1830.008 4.87 8.41 2007a, 2007b), but none of the techniques was developed to
BIA 2-059 0.2 0.2180.004 1.66 8.90 separate simultaneously all the analytes considered herein. Tak-
OXC 0.2 0.1850.011 6.35 7.65 ing into account the column manufacturer's recommendations
BIA 2-024 0.2 0.2280.019 8.61 14.3 for the use of methanol as the most powerful organic modifier,
Brain and also the chromatographic method previously reported by
BIA 2-265 0.06 0.0660.006 9.09 10.0 our research group (Alves et al., 2007a), the first experiments
S-Lic 0.06 0.0560.003 5.48 6.67 were performed at 30 C using a mixture of water and methanol
R-Lic 0.06 0.0660.002 2.40 9.94 (88:12, v/v) as mobile phase and a flow-rate of 0.7mL/min. Under
BIA 2-059 0.06 0.0640.005 7.64 7.15 these conditions, BIA 2–059, OXC and BIA 2–024 did not sepa-
OXC 0.06 0.0580.004 6.90 3.33 rate adequately and the percentage of methanol had to be di-
BIA 2-024 0.06 0.0640.003 5.30 7.30 minished. Thus a mobile phase composed of water–methanol
(92:8, v/v) pumped at 0.7ml/min was demonstrated to best
CNominal, Nominal concentration; CExperimental, experimental
correlate retention times and resolution of the peaks. In addition
concentration.
a it was also decided to evaluate the effects of temperature in the
Meanstandard deviation.
technique, as it can have a huge impact on peak resolution. The
effects of temperature ranging from 30 to 45 C on the resolution
of S-Lic, R-Lic, BIA 2–059, OXC and BIA 2–024 are shown in Fig. 2.
The chromatographic resolution was determined according to
phases of DDD, the major characteristic of this bioanalytical as- the method of Budakova et al. (2008), who recommend a resolu-
say is to provide sample analysis quickly, with high accuracy, tion factor higher than 1.5 as this corresponds to greater than
precision and selectivity. Moreover, considering that some of 90% separation between two peaks. As can be seen, the resolu-
the compounds under study are metabolites of each other, the tion of BIA 2–059 with OXC and OXC with BIA 2–024 was highest
analytical method has to be able to accurately distinguish all of at 30 and 35 C. Considering that at 30 C the run time was sub-
them. It was preferred to use a unique sample preparation and stantially longer than at 35 C, it was decided to use 35 C. These
analysis technique for all the compounds in both matrices, simple and economical chromatographic conditions with a
388

wileyonlinelibrary.com/journal/bmc Copyright © 2011 John Wiley & Sons, Ltd. Biomed. Chromatogr. 2012; 26: 384–392
Determination of eslicarbazepine and derivatives in mouse

Table 3. Intra- and inter-day accuracy (expressed as a percentage of nominal concentration) and precision (expressed as coeffi-
cient of variation, CV %) for BIA 2–265, eslicarbazepine (S-licarbazepine, S-Lic), R-licarbazepine (R-Lic), BIA 2–059, oxcarbazepine
(OXC) and BIA 2–024 in mouse plasma and brain homogenate quality control samples (n=5)

Matrix/ CNominal Intra-day Inter-day


drug (mg/mL) a
CExperimental (mg/mL) CV (%) Bias (%) a
CExperimental (mg/mL) CV (%) Bias (%)
Plasma
BIA 2-265 0.5 0.4760.01 3.37 4.75 0.5010.02 4.56 0.31
30 29.840.96 3.23 0.50 30.331.69 5.59 1.11
54 54.561.33 2.43 1.04 53.491.36 2.55 0.95
S-Lic 0.5 0.5310.02 4.41 6.19 0.4780.02 4.68 4.41
15 15.960.20 1.24 6.41 15.990.11 0.67 6.57
27 29.100.27 0.94 7.79 29.200.80 2.73 8.17
R-Lic 0.5 0.5120.02 4.57 2.33 0.5100.02 3.47 1.90
15 15.690.27 1.70 4.59 15.550.10 0.62 3.68
27 27.940.33 1.19 3.50 28.941.11 3.85 7.17
BIA 2-059 0.5 0.5210.02 4.84 4.14 0.5290.02 4.48 5.80
30 31.270.61 1.96 4.23 31.880.36 1.14 6.26
54 56.611.32 2.33 4.84 58.062.14 3.70 7.53
OXC 0.5 0.5060.01 1.93 1.13 0.4840.10 2.80 3.19
15 15.660.29 1.86 4.40 15.420.29 1.83 2.79
27 27.700.89 3.66 2.61 26.420.97 3.21 2.14
BIA 2-024 0.5 0.4860.02 3.92 2.87 0.4990.01 2.81 0.29
30 29.730.76 2.56 0.90 30.240.38 1.25 0.78
54 54.961.70 3.09 1.77 56.542.83 5.01 4.70
Brain
BIA 2-265 0.15 0.1470.01 5.68 1.80 0.1600.01 5.82 6.54
7.5 7.4940.31 4.15 0.08 7.4750.25 3.88 0.33
13.5 14.250.78 5.50 5.53 14.200.48 3.38 5.15
S-Lic 0.15 0.1470.01 3.69 1.81 0.152.001 5.60 1.45
7.5 7.2200.32 4.39 3.73 6.9610.31 4.44 7.19
13.5 13.390.69 5.16 0.80 13.120.53 4.01 2.83
R-Lic 0.15 0.1420.00 3.23 5.34 0.1560.01 7.01 4.22
7.5 7.2390.33 4.55 3.48 7.3280.34 5.43 2.29
13.5 13.680.72 5.28 1.30 13.970.31 2.22 3.46
BIA 2-059 0.15 0.1460.00 2.25 2.70 0.1570.01 5.79 4.48
7.5 7.3010.30 4.10 2.65 7.3070.32 4.35 2.57
13.5 13.950.71 5.06 3.31 14.140.61 4.31 4.71
OXC 0.15 0.1460.01 5.48 2.83 0.1540.01 5.99 2.72
7.5 7.1820.22 3.11 4.24 7.0600.54 7.72 5.87
13.5 13.300.67 5.01 1.49 13.630.67 4.89 0.97
BIA 2-024 0.15 0.1520.01 4.99 1.17 0.1590.01 3.67 5.70
7.5 7.2230.17 2.38 3.69 7.3770.44 5.91 1.64
13.5 13.010.77 5.93 3.60 13.350.50 3.71 1.10
CNominal, Nominal concentration; CExperimental, experimental concentration.
a
Meanstandard deviation.

mobile phase essentially composed of water are major advan- performing the precipitation with methanol, acetonitrile and tri-
tages of this technique, especially when applied in the initial chloroacetic acid (20%). The use of methanol and acetonitrile
phases of DDD. was not possible as they had very low selectivity. Trichloroacetic
Unfortunately, LC instrumentation and conditions are not the acid (20%) guaranteed optimum sample clean-up; however its
only limiting factor in high-throughput analysis and sample acidic properties easily degraded BIA 2–265 and the IS. Thus,
preparation must also be considered. Indeed, because of the a SPE procedure was developed, testing first the conditions
complex nature of plasma and brain homogenate, direct sample reported by Alves et al. (2007a). As under these conditions recov-
injection could not be considered a rugged approach and a pre- eries of analytes and IS were low, new conditions in washing
treatment procedure to remove proteins and other potential steps were tried and the best ones were obtained by washing
matrices interferences was needed prior to LC analysis. Protein the loaded cartridges four times with 1mL of water–methanol
precipitation was the first extraction procedure tested, (90:10, v/v). No endogenous interferences at the retention times
389

Biomed. Chromatogr. 2012; 26: 384–392 Copyright © 2011 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/bmc
A. Fortuna et al.

of the analytes and IS were found with the SPE procedure, allow- namely the fact that it is faster, less cumbersome and gives bet-
ing an accurate and selective sample preparation. Moreover, this ter extraction yields than the liquid–liquid extraction procedures
sample pre-treatment presents other important advantageous, reported in the literature (Pienimäki et al., 1995; Levert et al., 2002),

Table 4. Recovery of BIA 2–265, eslicarbazepine (S-licarbazepine, S-Lic), R-licarbazepine (R-Lic), BIA 2–059, oxcarbazepine (OXC)
and BIA 2–024 from plasma and brain homogenate supernatant of mouse (n=5)

Matrix
Drug Plasma Brain
a
CNominal (mg/mL) Recovery (%) CV (%) CNominal (mg/mL) Recoverya (%) CV (%)
BIA 2-265 0.5 91.041.81 1.99 0.15 100.23.10 3.09
30 87.984.70 5.34 7.5 98.115.01 5.10
54 85.624.16 4.85 13.5 103.12.71 2.63
S-Lic 0.5 107.76.69 6.21 0.15 107.54.69 4.37
15 109.35.35 4.89 7.5 100.94.68 4.64
27 107.44.52 4.21 13.5 108.33.65 3.37
R-Lic 0.5 101.13.55 3.51 0.15 103.43.42 3.31
15 108.94.98 4.58 7.5 101.94.94 4.85
27 109.30.88 0.80 13.5 108.53.27 3.01
BIA 2-059 0.5 103.58.78 8.48 0.15 109.86.35 5.78
30 105.05.43 5.18 7.5 99.564.72 4.74
54 104.92.41 2.30 13.5 106.44.41 4.13
OXC 0.5 97.287.23 7.43 0.15 96.822.65 2.74
15 95.844.95 5.17 7.5 93.795.49 5.86
27 96.473.85 3.99 13.5 98.273.19 3.24
BIA 2-024 0.5 84.605.39 6.35 0.15 81.711.85 2.27
30 85.004.26 5.01 7.5 86.693.94 4.54
54 86.484.19 4.84 13.5 93.292.78 2.98
CNominal, Nominal concentration.
a
Meanstandard deviation.

Table 5. Stability (values in percentage) of BIA 2–265, eslicarbazepine (S-licarbazepine, S-Lic), R-licarbazepine (R-Lic), BIA 2–059,
oxcarbazepine (OXC) and BIA 2–024 under conditions mimicking the entire assay process in plasma and brain homogenate super-
natant of mouse: after 30days at 30 C, 24h at 4 C, 2h in plasma and 2h in brain homogenate supernatant at room temperature
(RT), three freeze–thaw cycles and in mobile phase at 4 C for 24h

BIA 2-265 S-Lic R-Lic BIA 2-059 OXC BIA 2-024


Plasma CNominal 0.5 54 0.5 27 0.5 27 0.5 54 0.5 27 0.5 54
(mg/mL)
30 C, 30days 86.8a 87.8a 95.5 89.0 98.3 95.7 92.8a 88.8a 85.4 89.9 89.4 96.2

4 C, 24h 87.4 91.6 106.5 98.4 90.7 95.3 86.3a 85.7a 88.7 85.8 97.9 100.7
RT; 2h 87.0a 85.4a 95.2 97.1 92.1 109.7 86.1a 87.8a 85.2 90.1 89.3 103.1
Three freeze–thaw 89.4a 87.0a 93.3 103.2 88.0 89.1 87.3a 90.1a 102.2 99.8 108.9 108.1
cycles
Mobile phase 4 C, 95.0 98.8 103.9 99.9 96.5 98.5 99.2 102.3 99.7 102.2 94.0 101.5
24h
Brain CNominal 0.15 13.5 0.15 13.5 0.15 13.5 0.15 1.35 0.30 13.5 0.30 13.5
(mg/mL)
30 C, 30days 100.8 97.9 109.5 98.2 109.7 99.2 105.8 96.8 106.6 94.5 106.6 98.3

4 C, 24h 96.1 91.6 93.2 88.8 93.4 92.3 88.8 86.5 93.8 88.9 96.7 86.3
RT; 2h 90.9 92.9 93.9 93.2 92.1 95.2 87.4 88.5 89.5 85.9 90.6 90.9
Three freeze–thaw 99.2 101.6 102.6 99.4 100.3 101.3 96.3 96.1 95.9 96.2 94.3 91.1
cycles
Mobile phase 4 C, 101.6 101.1 103.5 100.5 106.6 100.5 103.9 99.8 100.4 100.4 100.7 94.4
24h
CNominal, Nominal concentration.
a
Results found at 30 C for 15days; at 4 C for 8h; at RT for 1h; and after two freeze–thaw cycles.
390

wileyonlinelibrary.com/journal/bmc Copyright © 2011 John Wiley & Sons, Ltd. Biomed. Chromatogr. 2012; 26: 384–392
Determination of eslicarbazepine and derivatives in mouse

and it is also suitable for automation with an on-line SPE and and S-Lic found in plasma and brain homogenates after their ad-
injection device. ministration to mice (Hainzl et al., 2001; Alves et al., 2008b) are
Finally, as we intended to use UV detection as it is the most likely to be determined using our methodology, because all con-
usual detection system, several wavelengths were analyzed centrations are included in our calibration ranges. Regarding BIA
and 235nm was chosen as the best compromise in terms of sen- 2–059 and BIA 2–024, only two in vivo studies have been per-
sitivity and selectivity. formed in order to trace their pharmacokinetic profiles after
drug administration to mice (Hainzl et al., 2001, 2002; Ambrósio
et al., 2002). Once again, our calibration ranges include the brain
Method Validation
and plasma concentrations found in vivo. As no in vivo studies
Considering the internationally accepted guidelines (Shah et al., have been performed for BIA 2–265 and it is structurally related
2000; US DHHS et al., 2001), two independent validation to the other compounds, the authors defined its calibration
studies were carried out for plasma and brain homogenate su- range considering those of the remaining analytes, using the
pernatant of mouse. The selectivity of the developed method largest ranges.
was demonstrated by representative chromatograms from Table 3 depicts the intra- and inter-day precision and accuracy
extracts of drug-free plasma and brain homogenates of mice of each of QC samples (QC1, QC2 and QC3) in mouse plasma and
and those from the correspondent extracts of samples spiked brain homogenates. Considering the validation performed for
with OXC, S-Lic, R-Lic, BIA 2–059, BIA 2–024, BIA 2–265 and IS plasma samples, the intra- and inter-day precision did not ex-
(Fig. 3). No interferences exist at the retention times of analytes, ceed 5.59% and the accuracy varied from 4.75 to 8.17%, while
showing the ability of the method to measure the compounds for brain homogenate, the intra- and inter-day precision did not
unequivocally in the corresponding biological matrices. The exceed 7.72% and the accuracy varied from 7.19 to 6.54%.
run time was 30min and the order of elution was BIA 2–265, These data demonstrate that both precision and accuracy were
S-Lic, R-Lic, BIA 2–059, OXC, BIA 2–024 and IS. within the acceptable limits in all matrices, being the method
The peak height ratios of analyte with IS in mouse plasma and precise and accurate to determine these drugs in mouse plasma
brain homogenate were linear in the concentration ranges used; and brain samples.
however the variance was not constant across the calibration The relative drug recoveries from plasma and tissue homoge-
range (heteroscedasticity was found) and hence a weighting fac- nates were estimated and are presented in Table 4. The mean
tor was used to improve residuals and to counter the variance relative recoveries, taking the compounds into account in all ma-
(Almeida et al., 2002). The best fit for the calibration curve was trices, ranged from 81.71 to 109.8%. The CV values were also de-
achieved by a linear equation of y=mx+c with 1/x2 as weighing termined and demonstrated to be low (≤ 8.48%), suggesting
factor. The calibration parameters for each compound, including consistent average recoveries over the evaluated concentration
slope, interception and regression coefficient (r 2) are shown in range. In plasma, the absolute recovery of the IS was 73.7% with
Table 1. Good linearity was found for all analytes in both matri- a CV value of 2.77%, while in brain homogenate it was found to
ces (r 2 ≥0.996). In addition, for all the compounds, LOQ values be 84.6% with a CV of 4.46%.
were established at 0.2 and 0.06mg/mL in plasma and brain From the drug stability data obtained in mouse brain homog-
homogenate, respectively. By evaluating five samples of blank enate under the conditions stated in the previous section, all the
plasma and blank brain homogenate supernatant spiked with compounds were demonstrated to be stable with no significant
each analyte at those concentrations, the values estimated for losses (Table 5). Considering the stability results obtained in
bias and CV were within the acceptance criteria (Table 2). At this plasma, also presented in Table 5, all the compounds were sta-
point it is important to highlight that the LOQ and the upper ble at all studied conditions except BIA 2–265 and BIA 2–059. In-
limits of the concentration ranges were defined considering deed, BIA 2–265 and BIA 2–059 were stable at 30 C for only 15
not only the instrumental capabilities and pre-validation results days, at room temperature for 1h, and for only two freeze–thaw
but also the previous concentrations found for the analytes in cycles. BIA 2–059 was also stable in plasma for only 8h at 4 C. Al-
some in vivo studies. There are several techniques for quantifi- though these compounds are less stable, this will have no im-
cation of OXC and/or racemic licarbazepine in biological ma- pact on application if the stability results are taking into account.
trices, but few of them are able to distinguish S-Lic and R-Lic
(Flesch et al., 1992; Volosov et al., 2000a, 2000b; Hainzl et al.,
2001; Alves et al., 2007a, 2007b; Mazzucchelli et al., 2007).
Conclusion
Among these, only two are able to simultaneously separate and A rapid and sensitive chiral LC-UV method was developed and vali-
quantify OXC, S-Lic and R-Lic in human plasma (Alves et al., dated for the simultaneous determination of OXC, S-Lic, R-Lic, BIA
2007b; Mazzucchelli et al., 2007) and only one in the plasma 2–059, BIA 2–024 and BIA 2–265 in the matrices of mouse
and brain of mouse (Alves et al., 2007b). In comparison to the plasma and brain homogenate. The adequate selectivity, sensi-
technique developed by Mazzucchelli et al. (2007), the LOQ tivity, precision and accuracy, with high recovery yields and ap-
values reported herein are higher probably because we based propriate run time, make it suitable for pharmacokinetic
our method on UV detection at 235nm instead of 210nm as studies and it will be easy to change from one compound to an-
Mazzucchelli et al. (2007). Although the use of shorter wave- other. In addition, because of the simple LC conditions and
lengths allows a lower LOQ for OXC, S-Lic and R-Lic, this was straightforward sample pre-treatment procedure, the method
not possible in our method because endogenous interferences is appropriate for application to elucidate the metabolic path-
were detected. The choice of 235nm was a compromise to ob- ways and pharmacokinetics of BIA 2–024, BIA 2–059 and BIA
tain good selectivity for all analytes evaluated. However, the 2–265. In addition, such pharmacokinetics data will be a useful
LOQ of the analytes obtained with the method developed was contribution to the development of an in silico approach that
often lower than those previously reported (Alves et al., 2007a, helps to understand the reasons for the success or failure of
2007b). On the other hand, the concentrations of OXC, R-Lic dibenz[b,f]azepine-5-carboxamide compounds.
391

Biomed. Chromatogr. 2012; 26: 384–392 Copyright © 2011 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/bmc
A. Fortuna et al.

Lipinski CA. Drug-like properties and the causes of poor solubility and
Acknowledgements poor permeability. Journal of Pharmacological and Toxicological Meth-
The authors would like to thank the Fundação para a Ciência e a ods 2000; 44: 235–249.
Tecnologia (SFRH/BD/31390/2006), Portugal and BIAL, Portugal. Luszczki JJ. Third-generation antiepileptic drugs: mechanisms of action,
pharmacokinetics and interactions. Pharmacological Reports 2009;
61: 197–216.
References Mazzucchelli I, Franco V, Marchiselli R, Perucca E and Gatti G. A novel
enantioselective microassay for the high-performance liquid chroma-
Almeida AM, Castel-Branco MM and Falcão AC. Linear regression for cal- tography determination of oxcarbazepine and its active metabolite
ibration lines revisited: weighting schemes for bioanalytical methods. monohydroxycarbazepine in human plasma. Therapeutic Drug Moni-
Journal of Chromatography B 2002; 774: 215–222. toring 2007; 29: 319–324.
Alves G, Figueiredo I, Castel-Branco M, Loureiro A, Falcão A and Nef P. Key animal models for the identification and validation of drug tar-
Caramona M. Simultaneous and enantioselective liquid chromato- gets. Drug Discovery Today 2001; 6: S91–S96.
graphic determination of eslicarbazepine acetate, S-licarbazepine, R- Neels HM, Sierens AC, Naelaerts K, Scharpé SL, Hatfiels GM and Lambert WE.
licarbazepine and oxcarbazepine in mouse tissue samples using ul- Therapeutic drug monitoring of old and newer anti-epileptic drugs.
traviolet detection. Analytica Chimica Acta 2007a; 596: 132–140; doi: Clinical Chemistry and Laboratory Medicine 2004; 42: 1228–1255; doi:
10.1016/j.aca.2007.05.056 10.1515/CCLM.2004.245
Alves G, Figueiredo I, Castel-Branco M, Loureiro A, Fortuna A, Falcão A Parrot N, Paquereau N, Coassolo P and Lavé T. An evaluation of the utility of
and Caramona M. Enantioselective HPLC-UV method for determina- physiologically based models of pharmacokinetics in early drug dis-
tion of eslicarbazepine acetate (BIA 2–093) and its metabolites in hu- covery. Journal of Pharmaceutical Sciences 2005; 94: 2327–2342; doi:
man plasma. Biomedical Chromatography 2007b; 21: 1127–1134; doi: 10.1002/jps.20419
10.1002/bmc.858 Pauli I, Timmers LFSM, Caceres RA, Soares MBP and Azevedo WF. In silico
Alves G, Fortuna A and Falcão A. High-performance liquid chromatography and in vitro: identifying new drugs. Current Drug Targets 2008; 9:
and its impact in the development of chiral drugs: a review. Trends 1054–1061.
in Chromatography 2008a; 4: 1–10. Pienimäki P, Fuchs S, Isojärvi J and Vähäkangas K. Improved detection and
Alves G, Figueiredo I, Falcão A, Castel-Branco M, Caramona M and determination of carbamazepine and oxcarbazepine and their meta-
Soares-da-Silva P. Stereoselective disposition of S- and R-licarbazepine bolites by high-performance liquid chromatography. Journal of Chro-
in mice. Chirality 2008b; 20: 796–804; doi: 10.1002/chir.20546 matography B 1995; 673: 97–105.
Ambrósio AF, Soares-da-Silva P, Carvalho CM and Carvalho AP. Mechanisms Schmitt W and Willmann S. Physiology-based pharmacokinetic modeling:
of action of carbamazepine and its derivatives, oxcarbazepine, BIA ready to be used. Drug Discovery Today: Technologies 2004; 1:
2–093, and BIA 2–024. Neurochemical Research 2002; 27: 121–130. 449–456; doi: 10.1016/j.ddtec.2004.09.006
Bohets H, Annaert P, Mannens G, Beijsterveldt L, Anciaux K, Verboven P, Shah VP, Midha KK, Findlay JWA, Hill HM, Hulse JD, McGilveray IJ, McKay
Meuldermans W and Lavrijsen K. Strategies for absorption screening G, Miller KJ, Patnaik RN, Powell ML, Tonelli A, Viswanathan CT and
in drug discovery and development. Current Topics in Medicinal Yacobi A. Bioanalytical method validation – a revisit with a decade
Chemistry 2001; 1: 367–383. of progress. Pharmaceutical Research 2000; 17: 1551–1557.
Budakova L, Brozmanova H, Grundmann M and Fischer J. Simultaneous Sui X, Sun J, Wu X, Li H, Liu J and He Z. Predicting the volume of distri-
determination of antiepelieptic drugs and their two active metabolites bution of drugs in humans. Current Drug Metabolism 2008; 9:
by HPLC. Journal of Separation Science 2008; 31: 1–8. 574–580.
Caldwell GW, Ritchie DM, Masucci JA, Hageman W and Yan Z. The new Tsaioun K, Bottlaender M and Mabondzo A. ADDME – avoiding drug devel-
pre-preclinical paradigm: compound optimization in early and late opment mistakes early: central nervous system drug discovery per-
phase drug discovery. Current Topics in Medicinal Chemistry 2001; spective. BMC Neurology 2009; 9: S1–S11; doi: 10.1186/1471-2377-9-
1: 353–366. S1-S1
Carlson TJ and Fisher MB. Recent advances in high throughput screening US DHHS, FDA, CDER. Guidance for Industry: Bioanalytical Method Valida-
for ADME properties. Combinatorial Chemistry and High Throughput tion. US Department of Health and Human Services, Food and Drug
Screening 2008; 11: 258–264. Administration, Center for Drug Evaluation and Research and Center
Flesch G, Francotte E, Hell F and Degen PH. Determination of the R-( ) for Veterinary Medicine, 2001. Available at: http://www/fda.gov/cder/
and S-(+) enantiomers of the monohydroxylated metabolite of oxcar- guidance/index.htm
bazepine in human plasma by enantioselective high-performance liq- Volosov A, Yagen B and Bialer M. Comparative stereoselective pharmaco-
uid chromatography. Journal of Chromatography 1992; 581: 147–151. kinetic analysis of 10- hydroxycarbazepine after oral administration of
Hainzl D, Parada A, and Soares-da-Silva P. Metabolism of two new antiepilep- its individual enantiomers and the racemic mixture to dogs. Epilepsia
tic drugs and their principal metabolites S(+)- and R( )-10,11-dihy- 2000a; 41: 1107–1111.
dro-10-hydroxy carbamazepine. Epilepsy Research 2001; 44: 197–206. Volosov A, Bialer M, Xiaodong S, Perucca E, Sintov A and Yagen B. Simulta-
Hainzl D, Loureiro AI, Parada A and Soares-da-Silva P. Metabolism of 10,11- neous stereoselective high-performance liquid chromatographic de-
dihydro-10-hydroxyimino- 5H-dibenz/b, f/azepine-5-carboxamide, a termination of 10-hydroxycarbazepine and its metabolite
potent anti-epileptic drug. Xenobiotica 2002; 32: 131–140; doi: carbamazepine-10,11-trans-dihydrodiol in human urine. Journal of
10.1080/00498250110087960 Chromatography B 2000b; 738: 419–425.
Levert H, Odou P and Robert H. Simultaneous determination of four antiepi- Yengi LG, Leung L and Kao J. The evolving role of drug metabolism in
leptic drugs in serum by high-performance liquid chromatography. drug discovery and development. Pharmaceutical Research 2007;
Biomedical Chromatography 2002; 16: 19–24; doi: 10.1002/bmc.117 24: 842–858; doi: 10.1007/s11095-006-9217-9
392

wileyonlinelibrary.com/journal/bmc Copyright © 2011 John Wiley & Sons, Ltd. Biomed. Chromatogr. 2012; 26: 384–392

You might also like