pharmacodynamic investigations of serotonergic agents

pharmacodynamic investigations of serotonergic agents

Journal of Pharmacological and Toxicological Methods 55 (2007) 214 – 223 Original article An integrated microdialy...

339KB Sizes 1 Downloads 129 Views

Journal of Pharmacological and Toxicological Methods 55 (2007) 214 – 223

Original article

An integrated microdialysis rat model for multiple pharmacokinetic/ pharmacodynamic investigations of serotonergic agents Christoffer Bundgaard a,⁎, Martin Jørgensen b , Arne Mørk c a


Discovery ADME, H. Lundbeck A/S, Copenhagen, Denmark Early Development Pharmacokinetics, H. Lundbeck A/S, Copenhagen, Denmark c Neurobiology, H. Lundbeck A/S, Copenhagen, Denmark Received 17 May 2006; accepted 31 July 2006

Abstract Introduction: Integrated in vivo models applying intracerebral microdialysis in conjunction with automated serial blood sampling in conscious, freely moving rodents are an attractive approach for pharmacokinetic (PK) and simultaneous pharmacokinetic/pharmacodynamic (PK/PD) investigations of CNS active drugs within the same animal. In this work, the ability to obtain and correlate data in this manner was evaluated for the selective serotonin (5-HT) reuptake inhibitor (SSRI) escitalopram. Methods: An instrumented rat model equipped with an intracerebral hippocampal microdialysis probe and indwelling arterial and venous catheters was applied in the studies. Concomitant with brain microdialysis, serial blood sampling was conducted by means of an automated blood sampling device. The feasibility of the rat model for simultaneous PK/PD investigations was examined by monitoring plasma and brain extracellular concentrations of escitalopram along with SSRI-associated pharmacological activity, monitored as changes in brain 5-HT levels and plasma corticosterone levels. Results: Combining intracerebral microdialysis and automated blood sampling did not cause any detectable physiological changes with respect to basal levels of plasma corticosterone or brain 5-HT levels. Furthermore, the PK of escitalopram in hippocampus following intravenous injection was not influenced by the presence of vascular catheters. Conversion of escitalopram dialysate concentrations into absolute extracellular levels by means of in vivo retrodialysis was verified by the no-net-flux method, which gave similar recovery estimates. The PK of escitalopram could be characterized simultaneously in plasma and the hippocampus of conscious, freely moving rats. Concomitantly, the modulatory and functional effects of escitalopram could be monitored as increases in brain 5-HT and plasma corticosterone levels following drug administration. Discussion: The applicability of intracerebral microdialysis combined with arterial blood sampling was demonstrated for simultaneous PK/PD investigations of escitalopram in individual rats under non-stressful conditions. Together, these temporal relationships provide multiple PK/PD information in individual animals, hence minimizing inter-animal variation using a reduced number of animals. © 2006 Elsevier Inc. All rights reserved. Keywords: 5-HT (5-Hydroxytryptamine, serotonin); Automated blood sampling; Corticosterone; Escitalopram; Methods; Microdialysis; PK/PD; Rat; SSRI

1. Introduction Preclinical pharmacokinetic (PK) characterization and in vivo pharmacological properties of new chemical entities are important components during lead compound selection, optimization and elucidative mechanistic investigations in vivo during the drug discovery process. Accordingly, reliable tech⁎ Corresponding author. H. Lundbeck A/S, 9 Ottiliavej, DK-2500 Copenhagen-Valby, Denmark. Tel.: +45 3630 1311; fax: +45 3643 8302. E-mail address: [email protected] (C. Bundgaard). 1056-8719/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.vascn.2006.07.001

niques are needed that can generate the requisite pharmacokinetic/pharmacodynamic (PK/PD) information for an increased number of compounds. When dealing with compounds targeting the central nervous system (CNS), the distributional behaviours both peripherally (blood) and centrally (brain) are important determinants of in vivo drug effects. Several techniques have been employed for the study of drug transport to and distribution within the brain (Pardridge, 1998). Of these, intracerebral microdialysis has gained growing popularity and owing to the quantitative developments, this technique has been used increasingly over the past years for in vivo sampling of

C. Bundgaard et al. / Journal of Pharmacological and Toxicological Methods 55 (2007) 214–223

extracellular exogenous compounds in brains of freely moving rats (for a recent review see Plock & Kloft, 2005). The free extracellular brain concentration measured by microdialysis may reflect the amount of drug available at the pharmacological target for e.g. receptor or transporter occupancy. As the microdialysis technique also allows the simultaneous determination of different endogenous substances, such as neurotransmitters, in the same local interstitial environment, it represents an attractive tool for PK/PD investigations of CNS active drugs (de Lange, Ravenstijn, Groenendaal, & van Steeg, 2005). With the recent advent of automated blood sampling systems designed for small serial blood removals from catheterized conscious, freely moving rats (Gunaratna, Kissinger, Kissinger, & Gitzen, 2004; Wang, Ho, Yen, & Tsai, 2006), integrated animal models combining microdialysis and blood sampling can be established that permit simultaneous collection of both biochemical data and drug concentrations in different biofluids in individual animals with minimal human intervention. Such animal models may improve both the accuracy and quality of data by minimizing the inter-animal variation using a reduced number of animals. In addition, stress caused by the use of conscious but restrained animals may affect the measured pharmacological parameters, as well as the metabolism of the drug under investigation (Shimizu, Take, Hori, & Oomura, 1992; Vallée, Mayo, Maccari, Le Moal, & Simon, 1996; Watanabe et al., 2002). The use of anesthetized animals may also interfere with the pharmacological activity and may also alter the PK of the drug by reducing hepatic cytochrome P450 activity or by decreasing renal function (Gumbleton & Benet, 1991; Loch, Potter, & Bachmann, 1995). Chronically instrumented freely moving animals equipped with indwelling vascular catheters or intracerebral microdialysis probes may, however, also be subjected to stress caused by the surgical instrumentation (Drijfhout et al., 1995; Fagin, Shinsako, & Dallman, 1983). Furthermore, from a PK/PD perspective, appropriate estimation of recovery of the drug under investigation through the probe membrane is a prerequisite in order to convert dialysate levels into true extracellular drug concentrations. Therefore, such animal models should be properly validated with respect to the intended PK and pharmacological measurements, as well as to animal welfare considerations. This is particularly important when levels of endogenous biochemical markers susceptible to various stressful stimuli (such as neurotransmitters or stress hormones) are applied as pharmacological endpoints in PK/PD investigations. For serotonergic agents such as antidepressants, these endpoints are frequently employed as quantitative measures of the pharmacological effect. Thus, the objective of the present study was to assess the applicability of intracerebral microdialysis in conjunction with automated serial blood sampling as a tool for simultaneous PK/PD characterization of the selective serotonin reuptake inhibitor (SSRI) escitalopram. Escitalopram is the S(+)-enantiomer of the racemic compound citalopram and is the isomer responsible for the antidepressant efficacy (Sánchez, Bøgesø, Ebert, Reines, & Bræstrup, 2004; Thase, 2006).


2. Methods 2.1. Chemicals Escitalopram and a chlorine substituted internal standard (Lu 10-202) were provided by H. Lundbeck, A/S, Denmark. Escitalopram was used as oxalate salt dissolved in 0.9% NaCl. 5-HT creatinine sulfate, corticosterone, dexamethasone, and decolorizing carbon (activated charcoal Norit®) were obtained from Sigma–Aldrich (St. Louis, MO, USA). 2.2. Animals and surgical procedures Male Sprague–Dawley rats (Charles River Lab., UK) about 10 weeks of age (weighing 310–400 g) were used in these studies. Rats were acclimatised for at least 5 days prior to surgery under a 12 h light/dark cycle (lights on at 06:00 h) and had free access to water and a standard rodent diet. Ethical permission for the procedures used in these studies was granted by the animal welfare committee, appointed by the Danish Ministry of Justice and all animal procedures were carried out in compliance with EC Directive 86/609/ EEC and with the Danish law regulating experiments on animals. Surgical preparation of the animals was done in two phases, each followed by a recovery period of 48 h. Implantation of the intracerebral microdialysis probe guide cannula was conducted in the first phase, followed by vascular catheterization in the second phase. Anaesthesia was induced by a mixture of Hypnorm® (fentanyl 0.32 mg/ml-fluanisone 10 mg/ml, Janssen, UK), Dormicum® (midazolam 5 mg/ml, Roche, Switzerland) and water (1:1:2 v/v) (0.25 ml/100 g body weight). Body temperature was maintained at 37 °C using a thermostated heating pad (CMA/150, CMA, Sweden). The intracerebral guide cannula (CMA/12, CMA) was stereotaxically implanted into the brain, aiming to position the dialysis probe tip in the ventral hippocampus (coordinates: 5.6 mm posterior to bregma, lateral − 5.0 mm, 7.0 mm ventral to dura according to the coordinates of a rat brain atlas (Paxinos & Watson, 1986)). Anchor screws and acrylic cement were used for fixation of the guide cannulas. Following recovery, indwelling pyrogen-free catheters (Tygon®, 0.4 mm i.d.; 0.79 mm o.d., Datainnovation AB, Sweden) were implanted in the external jugular vein and in the common carotid artery according to standard surgical procedures (van Dongen, Remie, Rensema, & van Wunnik, 1990). In some experiments, rats were only cannulated in the jugular vein. After cannulation, the catheters were externalized subcutaneously and exited at the back of the neck through a mesh button fastened to the back muscles and tunneled through a protective metal spring anchored to the button. Rats were attached to a dual channel swivel (Instech 375/D/22, Instech Lab., PA, USA) and the arterial catheter was continuously infused with a heparinized saline solution (20 IU/ml, 3.5 μl/min) in order to maintain catheter patency. Food and water intake were frequently monitored during recovery.


C. Bundgaard et al. / Journal of Pharmacological and Toxicological Methods 55 (2007) 214–223

2.3. Experimental protocols 2.3.1. General brain microdialysis and blood sampling procedures On the day of experiment, a microdialysis probe (CMA/12; o. d. 0.5 mm, membrane length 3 mm, CMA) was lowered into the hippocampus of the conscious rat. Perfusion of the microdialysis probe with filtered artificial cerebrospinal fluid (aCSF) (145 mM NaCl, 3 mM KCl, 1 mM MgCl2, 1.2 mM CaCl2, pH 7.4) was begun shortly before probe insertion and continued for the duration of the experiment delivered by a microinjection pump at a constant flow-rate of 2 μl/min, unless otherwise stated. The probe outlet tubing was passed to a collection vial attached to the swivel, allowing the animal to move freely in its cage during the study. A 3.5 h stabilization period was allowed following probe insertion after which time microdialysis sampling (20 μl fractions) was initiated. 5-HT samples were collected in vials containing 5 μl 25 mM ascorbic acid and 1.35 mM EDTA to prevent degradation (Thorré, Pravda, Sarre, Ebinger, & Michotte, 1997). Concomitant with microdialysis sampling, serial arterial blood sampling was accomplished by an automated blood sampling device (AccuSampler®, Datainnovation AB), consisting of a robotic PC-controlled sampling system with an embedded refrigerated fraction collector. Blood samples were withdrawn at predefined time intervals and every sample was automatically replaced by an equal volume of sterile heparinized saline to maintain the same circulatory volume in the rat. The total volume of blood collected during each experiment did not exceed 15% of the total blood volume (Diehl et al., 2001). After the experiments the animals were sacrificed, their brains were removed, frozen and sliced for verification of microdialysis probe placement. Blood samples were centrifuged at 6500 ×g for 10 min at 4 °C, and plasma was harvested and frozen until analysis. All experiments were performed between 08:00 h and 17:00 h. 2.3.2. Microdialysis recovery determinations In vitro recovery of escitalopram was determined as recovery gained from the surrounding media of the probe, and compared to recovery by loss from the probe (retrodialysis). Probes were immersed in aCSF solutions containing 50, 100 and 250 ng/ml escitalopram and perfused with drug-free aCSF at 2 μl/min. In vitro recovery by gain (Recgain) was calculated using the expression: Recgain ð%Þ ¼

Cdial  100 Cm


where Cdial is the average concentration in the dialysate samples and Cm is the concentration in the surrounding medium. Recovery by loss was examined by perfusing escitalopram solutions through the probes submerged in blank aCSF. Recovery by loss (Recloss) was calculated according to: Recloss ð%Þ ¼

Cin −Cout  100 Cin


where Cin is the concentration entering the probe and Cout is the average concentration leaving the probe.

In vivo recovery was estimated by retrodialysis (Bouw & Hammarlund-Udenaes, 1998) by perfusing the brain-implanted probe with 50 ng/ml escitalopram at a flow-rate of 2 μl/min. The fractional loss was measured following a 40 min equilibration period and recovery was calculated by Eq. (2). To assess the validity of in vivo retrodialysis, the method was compared to the no-net-flux method at steady-state (Lönnroth, Jansson, & Smith, 1987). Steady-state extracellular escitalopram concentrations of approximately 200 ng/ml were attained by administering an i.v. loading dose of 6 mg/kg followed by continuous infusion of 4 mg/kg/h, estimated via pilot studies. When steady-state was established, probes were perfused with successive escitalopram solutions (20–400 ng/ml) at 2 μl/min and the gain or loss from the perfusate was measured at each concentration level. Recovery was obtained by plotting (Cout − Cin) versus Cin for each individual animal. Applying linear regression, the slope of the fitted straight line yields the dialysis recovery. 2.3.3. Automated blood sampling recovery The potential drug loss due to dilution and/or adsorption inside the tube system during a sample cycle in the automated blood sampling device was assessed in vitro for escitalopram. Heparinized blood pools spiked with escitalopram in concentrations ranging from 20 to 500 ng/ml were used for these studies. Samples (200 μl) were siphoned through the sampling system and the amount of drug recovered after a sample cycle was compared to reference values obtained by manual pipetting from the same source. 2.3.4. Assessment of effects of surgical instrumentation The effect of catheterization of the carotid artery on the brain penetration and hippocampal PK of escitalopram was investigated by microdialysis. The test compound was administered as an i.v. bolus (5 mg/kg) to two groups of rats (n = 5 in each group), one of which had vascular catheters implanted in the carotid artery and in the jugular vein, while the other only had the jugular vein catheterized, and the brain concentration-time profiles were monitored for 3 h. The effects of surgical procedures on biochemical and endocrine parameters were evaluated based on basal brain 5-HT and plasma corticosterone levels. The basal 5-HT output in the hippocampus was assessed in two groups of rats with or without vascular cannulation (n = 5 in each group). 6 consecutive dialysis samples (20 μl) were collected from each rat at a perfusion flow-rate of 1 μl/min. The potential endocrine stress response following microdialysis probe implantation was studied by measuring circulating corticosterone levels in controls and probe implanted catheterized rats (n = 5 in each group). In addition, the plasma corticosterone time-course during repeated blood sampling was assessed as a marker of the stress intensity associated with the sampling procedure. 2.3.5. Pharmacokinetic and pharmacodynamic studies Rats received an i.v. bolus of escitalopram (2.5 mg/kg) or vehicle via the jugular catheter (5 rats per group). Arterial blood

C. Bundgaard et al. / Journal of Pharmacological and Toxicological Methods 55 (2007) 214–223

samples (200 μl) were drawn at regular time intervals for drug and corticosterone analysis. In conjunction, microdialysis was performed at a flow-rate of 2 μl applying a 10-minute sampling regimen, resulting in a temporal resolution of 20 min, as every second dialysis sample was analysed for escitalopram and 5-HT. Time points were corrected for lag time of the perfusate from the microdialysis site to the probe outlet. In vivo recovery estimation of escitalopram through the probe was conducted following wash out at the completion of each experiment by means of retrodialysis. 2.4. Bioanalytical assays 2.4.1. Analysis of escitalopram The plasma concentrations of escitalopram were determined by HPLC with fluorescence detection after a two-step liquid– liquid extraction procedure, modified from (Sidhu et al., 1997). After addition of the internal standard 50 μl plasma samples were alkalinized with 1 N NaOH, and extracted with 6 ml heptane containing 1.5% isoamyl alcohol by shaking for 15 min. Following centrifugation, the organic layer was mixed with 100 μl 0.1 N HCl and shaken for 15 min. After centrifugation, 70 μl of the aqueous phase was injected into the chromatographic system (L-7000 Merck-Hitachi, Germany). Separation was carried out at 30 °C on a Hypersil C18 column (200 mm × 4.6 mm i.d., 5 μm; Agilent, CA, USA). For the mobile phase, a 44 mM phosphate buffer containing 0.1% triethyl amine (pH 4.5) was prepared and mixed with acetonitrile in a ratio of 55:45 (v/v). The flow-rate was set at 1.2 ml/min and fluorescence emission of the analytes was monitored at 296 nm, with excitation at 240 nm. Calibration curves were established in the range of 2–600 ng/ml. The limit of detection was 1.6 ng/ml (signal-to-noise (S/N) ratio of 3). For determination of escitalopram in brain microdialysates, 20 μl samples were injected directly onto a reverse phase narrow bore column (Xterra, 150 × 2.1 mm i.d., 3.5 μm, Waters, MA, USA) operated at a flow-rate of 0.2 ml/min followed by fluorescence detection. Elution was performed at 30 °C using a mobile phase comprising of 44 mM phosphate buffer containing 0.1% triethyl amine (pH 4.5): acetonitrile (71:29 v/v). Standard curves were constructed in the range of 1–60 ng/ml, prepared in aCSF. The minimal detectable concentration of escitalopram was 0.5 ng/ml (S/N = 3). 2.4.2. Analysis of corticosterone Plasma concentrations of corticosterone were assayed by liquid–liquid extraction followed by HPLC analysis with UVdetection. The assay was optimized to 50 μl plasma samples based on previously published methods (Wong, Chien, & D'mello, 1994; Woodward & Emery, 1987). Extraction was performed by adding internal standard (dexamethasone), 5 ml diethyl ether-dichloromethane (60:40 v/v) and 0.5 ml 3 M NaOH to 50 μl plasma. After shaking for 15 min, the mixture was centrifuged and the organic layer was washed with 1 ml deionized water, followed by recentrifugation and separation. The organic layer was evaporated to dryness under nitrogen steam at room temperature and the residue was redissolved in


70 μl water:acetonitrile (80:20 v/v). 60 μl was injected into the HPLC system (L-7000 Merck–Hitachi, Germany). Chromatography was performed at 30 °C on a Luna C18 column (100 × 4.6 mm i.d., 5 μm, Phenomenex, CA, USA). The mobile phase consisted of acetonitrile–acetic acid (83 mM, pH 3.0) (25:75 v/v) pumped at a flow-rate of 1.0 ml/min. The effluent was monitored at 254 nm. Calibration standards in the range of 20–400 ng/ml were prepared from pooled heparinized plasma, which had been stripped of all endogenous corticosterone with decolorizing carbon. Recovery was 92% and independent of the concentration used. Using this assay method, the limit of detection was 10 ng/ml (S/N = 3). 2.4.3. Analysis of 5-HT The concentration of 5-HT in the dialysates was determined by means of HPLC with electrochemical detection. 20 μl sample aliquots were injected into the chromatographic system by a refrigerated autosampler (4 °C) (Agilent 1100 series). Separation was achieved at 30 °C on a MD-150 C18 column (150 × 3.2 mm i.d., 3 μm particle size, ESA Inc., MA, USA) using a mobile phase consisting of 75 mM NaH2PO4 (pH 3.0), 25 μM EDTA, 0.1% triethyl amine and 9% (v/v) methanol delivered at a flow-rate of 0.4 ml/min. Electrochemical detection was accomplished using a Coulometric detector (Coulochem III, ESA Inc.) equipped with a dual electrode analytical cell (5014B Dual Channel Microdialysis cell, ESA Inc.). The first electrode was set at 0 mV and the second at 220 mV with respect to a palladium reference electrode. 5-HT calibration standards were prepared in aCSF from serotonin creatinine sulfate stock solutions. Based on a S/N ratio of 3:1, the detection limit for 5-HT was 1 fmol per 20 μl. 2.5. Data analysis and statistics Differences in recovery obtained by the different methods were assessed by Student's t-test. Concentration dependence was assessed by one-way analysis of variance (ANOVA). PK analysis of plasma and brain escitalopram concentration-time curves was performed individually according to compartmental modelling using WinNonlin™ ver. 4.1 (Pharsight Corp., CA, USA). Prior to PK analysis, brain extracellular escitalopram concentrations were calculated by correcting dialysate concentrations for in vivo recovery of the respective probes. 5-HT levels in microdialysates were presented either as fmol/ sample or converted to percentages of pre-dose levels. The mean value of three consecutive 5-HT samples immediately preceding drug injection served as the basal 5-HT level for each experiment (normalized to 100%). To assess drug/vehicle induced changes in extracellular 5-HT and plasma corticosterone levels across time, one-way ANOVA for repeated measures was used, followed by post hoc comparisons of individual means accomplished by Tukey's test. To determine differences between treatments, a two-way ANOVA was applied. Statistical tests were performed using SigmaStat™ ver. 3.0. (Systat Software Inc, CA, USA). All results were expressed as mean values (±SD) and statistical significance was defined as p b 0.05.


C. Bundgaard et al. / Journal of Pharmacological and Toxicological Methods 55 (2007) 214–223

3. Results 3.1. Evaluation of microdialysis and blood sampling efficiency The concept of in vivo retrodialysis assumes directional independence of the solute through the dialysis membrane. In order to verify this assumption, in vitro experiments were performed comparing recovery of escitalopram by gain from the surrounding drug medium to recovery determined by loss from the perfusate containing the compound. The results are summarized in Table 1. Based on one-way ANOVA, no differences in relative gain or loss were found between the three concentration levels tested in the medium or perfusate (Gain, F14,2 = 0.95, p = 0.42; Loss, F14,2 = 0.36, p = 0.70). In addition, escitalopram exhibited similar mean recovery determined by relative gain or loss (t-test, p = 0.53). These results indicate that there is no directional bias for escitalopram in crossing the dialysis membrane and points towards a minimal influence of adhesion of the drug to the dialysis membrane and tubing material. The method of in vivo retrodialysis was evaluated as a means to obtain absolute extracellular concentrations of escitalopram in the brain. Applying 50 ng/ml escitalopram in the perfusion fluid, relative recovery in ventral hippocampus was 14.2 ± 6.3% (n = 8). Recovery of escitalopram calculated by in vivo retrodialysis was similar to the results obtained in vitro (t-test, p = 0.81) indicating that there was a non-significant difference in diffusivities for escitalopram in water (aCSF) versus brain tissue. By continuous measurement of the loss of escitalopram from the perfusate during retrodialysis, in vivo recovery was shown to be relatively stable for up to 8 h and no time dependent trends could be detected (Fig. 1). In vivo retrodialysis was therefore considered reliable over the experimental period of a PK study. In vivo recovery of escitalopram obtained using the no-netflux method under steady-state was 15.9 ± 2.9% (n = 6) at a perfusion flow-rate of 2 μl/min. Recovery did not differ significantly from the results obtained by in vivo retrodialysis (t-test, p = 0.54). Under the conditions applied during the no-net-flux experiment, drug diffusion into or out of the brain was observed in all rats, depending on the concentration added to the perfusion fluid. A linear relationship between Cin and (Cout − Cin) was established in all animals tested. Linear regression performed separately above and below the intercept of the abscissa resulted in

Fig. 1. Stability of in vivo recovery of escitalopram determined by retrodialysis in three different rats. Probes were continuously perfused with 50 ng/ml escitalopram at 2 μl/min.

similar slopes on both sides of the intercept for each individual rat. This indicated that in vivo recovery was not affected by the direction of diffusion, confirming the results obtained in vitro. Fig. 2 shows no-net-flux plots and linear regressions obtained from three representative rats. The efficiency of the automated blood sampling system with respect to obtaining accurate drug concentrations for PK profiling of escitalopram was evaluated by quantitating the drug loss during a sample cycle. Under controlled in vitro conditions, escitalopram concentrations measured after automated sampling from spiked blood pools were compared to the corresponding references obtained manually (Table 2). The recovery of escitalopram following automated sampling averaged 85.3 ± 2.5%. A linear relationship was obtained between results from manual and automated sampling at the four concentration levels tested (r2 = 0.996, p b 0.001) demonstrating that sampling recovery was independent of the drug

Table 1 Comparison of mean in vitro recoveries of escitalopram determined by relative gain and loss Concentration (ng/ml)

Recovery (%) By gain (n = 5)

By loss (n = 5)

50 100 250 Average

15.1 ± 1.0 13.5 ± 2.0 13.1 ± 3.1 13.9 ± 2.3

13.2 ± 3.6 12.6 ± 2.2 14.3 ± 3.8 13.4 ± 3.1

CMA/12 probes with a 3 mm polycarbonate membrane were used, applying perfusion flow-rates of 2 μl/min.

Fig. 2. In vivo recovery determination of escitalopram using the no-net-flux method in three individual rats. Cin and Cout represents concentrations of escitalopram in the perfusion fluid and dialysate, respectively. The study was performed under steady-state achieved by continuous i.v. infusion of escitalopram before and during the experiment. Microdialysis perfusion flow-rate was 2 μl/min. The slope (α) signifies the relative recovery determined by linear regression in each experiment.

C. Bundgaard et al. / Journal of Pharmacological and Toxicological Methods 55 (2007) 214–223


Table 2 In vitro sampling recovery of escitalopram in the applied automated sampling device Spiked blood Measured plasma concentration (ng/ml) concentration Automated sampling Manual sampling (ng/ml) (n = 3) (n = 3)

Recovery (%)

20 100 250 500

88.5 83.2 85.7 83.7

6.9 ± 0.5 29.7 ± 3.0 66.6 ± 3.1 130.8 ± 4.2

7.8 ± 0.3 35.7 ± 0.6 77.7 ± 2.5 156.2 ± 2.9

Recovered amount is expressed as the average concentration found following three sampling cycles relative to samples withdrawn manually from the same source.

concentration used. Thus, in order to establish absolute plasma concentrations of escitalopram applying the automated sampling system in vivo, a correction factor of 100/85 was incorporated.

Fig. 4. Brain extracellular fluid levels of escitalopram in the hippocampus of artery catheterized (▴) and control rats (●) measured by microdialysis (1 μl/min) following an intravenous bolus dose of 5 mg/kg (n = 5 in each group). The levels of escitalopram were corrected for in vivo recovery by retrodialysis in each individual rat (averaged 28.8± 8.7%, n = 10, 1 μl/min).

3.2. Biochemical findings following surgical instrumentation Basal 5-HT microdialysate levels from the hippocampus of rats with implanted vascular catheters were compared to basal 5-HT levels in non-catheterized rats. Two days after surgery mean basal 5-HT levels were 7.3 ± 1.8 fmol/sample and 8.0 ± 1.9 fmol/sample, in the non-catheterized and catheterized group, respectively, indicating that the basal hippocampal 5-HT output was not influenced by the presence of catheters (t-test, p = 0.12). Daytime circulating plasma corticosterone levels in microdialysis probe-implanted rats and in controls are shown in Fig. 3. Mean basal corticosterone levels were 145.3 ± 52.1 ng/ml and 151.6 ± 42.4 ng/ml in controls and probe-implanted rats, respectively. No characteristic secretory pattern was observed and corticosterone levels remained relatively stable throughout the experimental period. Two-way ANOVA revealed that both treatment (with or without probe implanted) and time were nonsignificant (F1,69 = 0.27, p = 0.61 and F6,69 = 0.15, p = 0.98).

Fig. 3. Mean time profiles of circulating corticosterone levels in conscious rats in probe implanted rats (▵) and controls (▴) (n = 5 in each group). Samples (200 μl) were collected every 80 min through the artery catheter using an automated blood sampling device.

Hence, the presence of hippocampal microdialysis probes did not result in increased basal corticosterone levels compared to controls, indicating a lack of stress-induced hypothalamic–

Fig. 5. A: Mean (±SD) escitalopram plasma (total drug) (○) and brain extracellular (●) concentration-time plots after i.v. administration of 2.5 mg/kg (n = 5). B: Plasma and brain data from one representative rat including PK fits according to two-compartment models with first-order elimination. Microdialysis was performed at a flow-rate of 2 μl/min. Brain dialysate levels were converted to extracellular concentrations by in vivo recovery via retrodialysis in each individual rat. Plasma escitalopram levels were corrected for an estimated drug loss of 15% during transfer within the automated sampling system.


C. Bundgaard et al. / Journal of Pharmacological and Toxicological Methods 55 (2007) 214–223

pituitary–adrenal (HPA) activity caused by the presence of the probe. 3.3. Pharmacokinetic and pharmacodynamic findings To investigate a potential effect of artery catheterization on the blood–brain barrier (BBB) penetration and PK of escitalopram in the hippocampus, microdialysis was employed in rats with and without artery-implanted catheters. Following an i.v. injection of 5 mg/kg, mean peak levels (Cmax) of escitalopram were 152.1 ± 29.5 ng/ml and 173.1 ± 26.6 ng/ml for artery catheter-implanted rats and controls, respectively, indicating that the extent of BBB penetration was not affected by the presence of the catheter (t-test, p = 0.33) (Fig. 4). Neither was total escitalopram exposure, expressed as AUC, significantly different between the two groups (AUC0–170 = 13,573 ± 1873 ng × min/ml vs. 11,576 ± 2786 ng × min/ml, t-test, p = 0.28). Due to fast BBB penetration, the absorption rates of escitalopram into the brain could not be calculated with the current data. However, based on Cmax and AUC, these observations indicate that catheterization of the carotid artery did not influence the penetration rate and exposure of escitalopram in the brain extracellular fluid. The instrumented rat model was used to characterize the PK of escitalopram in brain and plasma simultaneously with serotonergic effects expressed as brain 5-HT output and changes in plasma corticosterone levels. Following an i.v. bolus of 2.5 mg/kg, escitalopram was rapidly distributed into the brain and declined parallel to plasma concentrations (Fig. 5). The main PK parameters are listed in Table 3. Considering that plasma protein binding of escitalopram amounts to ≈ 50% in rats (H. Lundbeck A/S, unpublished data), the brain/unbound plasma AUC ratio was estimated to be 0.9 ± 0.2 (n = 5). The extensive brain exposure was reflected by the high volume of distribution found for escitalopram in these experiments. The acute pharmacological response measured as hippocampal 5-HT output following the escitalopram bolus is depicted in Fig. 6A. Escitalopram induced a significant increase Table 3 Summary of mean pharmacokinetic parameter estimates of escitalopram in plasma and brain following administration of 2.5 mg/kg as i.v. bolus (n = 5) Mean ± SD Plasma AUC0−inf total (ng × min/ml) AUC0−inf unbound (ng × min/ml)# CL (ml/min/kg) Vdss (l/kg) t1/2,β (min) Hippocampus AUC0−inf unbound (ng × min/ml) t1/2,β (min) Brain/plasma AUC ratio, unbound

10546 ± 2995 5273 ± 1498 252 ± 67 19.8 ± 4.2 74⁎ ± 25 4631 ± 826 100⁎ ± 46 0.9 ± 0.2

Parameters were calculated based on two-compartment models for both plasma and brain concentration-time data fitted individually in each rat. AUC: area under curve; CL: clearance; Vdss: volume of distribution; t1/2,β: elimination halflife. ⁎ Not statistically different (t-test, p = 0.30). # Calculated based on estimated plasma protein binding of 50%.

Fig. 6. Effect of i.v. administration of escitalopram on extracellular levels of 5HT in ventral hippocampus (A) and plasma corticosterone levels (B). Escitalopram (2.5 mg/kg bolus, closed symbols; n = 5) or vehicle (open symbols, n = 5) was administered at t = 0 min. Microdialysis was performed at a flow-rate of 2 μl/min. ⁎p b 0.05.

in extracellular 5-HT relative to the pre-dose baseline level (F20,96 = 14.64, p b 0.001). Maximal 5-HT output was obtained in the first dialysis sample post drug administration and averaged 410 ± 101% of basal levels. Post-hoc analysis revealed that the 5-HT output remained significantly elevated for 160 min following drug administration. The functional consequence of increased extracellular 5-HT was assessed by measuring the stimulation of the HPA-axis expressed as changes in plasma corticosterone levels. As shown in Fig. 6B, escitalopram produced a transient, but significant increase in corticosterone levels across time (F12,59 = 6.89, p b 0.001). Hence, a significant effect was found from 10–40 min post drug administration based on post-hoc analysis. Vehicle treatment had no significant effect on extracellular 5-HT levels nor on plasma corticosterone levels (F20,88 = 0.94, p = 0.54; F12,63 = 0.65, p = 0.78, respectively). 4. Discussion In these studies, the performance of the intracerebral microdialysis technique applied in conjunction with automated

C. Bundgaard et al. / Journal of Pharmacological and Toxicological Methods 55 (2007) 214–223

serial blood sampling in freely moving conscious rats was evaluated with reference to simultaneous PK/PD evaluation of escitalopram. When applying microdialysis in the study of pharmacokinetics, a major concern lies in the accurate determination of in vivo recovery of the compound of interest. Retrodialysis is a simple method to obtain in vivo recovery estimates in each experimental animal. However, as the compound is delivered via the perfusate, this method assumes a directional independence of the solutes through the dialysis membrane. Since recovery is dependent on a number of factors, including physical–chemical properties of the compound, membrane characteristics and matrix tortuosity (Stenken, 1999), a directional bias has been shown for a number of compounds, compromising the usefulness of retrodialysis (Burgio & McNamara, 1993; Groth & Jørgensen, 1997; Schuck, Rinas, & Derendorf, 2004). Therefore, the method of retrodialysis was validated for escitalopram in vitro and in vivo. In the current studies, no directional dependence of diffusion of escitalopram across the dialysis membrane was found and in vivo retrodialysis resulted in similar recovery values to those obtained by using the no-net-flux method at steady-state. Furthermore, it was demonstrated that in vivo recovery determined by retrodialysis was relatively stable when assessed over an extended period. Together, these findings support the use of in vivo retrodialysis by drug to accurately facilitate conversion of dialysate concentrations of escitalopram into absolute extracellular fluid concentrations in the brain. The similar recovery estimates between the method of no-net-flux and retrodialysis found in vivo for escitalopram in the current study have also been shown for other compounds like the antiviral nucleoside zidovudine and different melatonin analogues (Fox et al., 2002; Le Quellec et al., 1995). However, for acetaminophen, caffeine and ibuprofen differences between the two methods have been reported (Song & Lunte, 1999; Tegeder et al., 1999). These findings may be related to differences in lipophilicities or changes in the biological system e.g. saturable active transport across the BBB. Thus, this underlines that applying intracerebral microdialysis for quantitative purposes, the probe calibration procedure should be appropriately validated especially when dealing with lipophilic substances or compounds subjected to active transport processes. When microdialysis sampling is applied for simultaneous PK/PD evaluations, the methodological aspects of the technique have to be considered in relation to the quantitative analysis and the desired temporal resolution. High dialysis recovery is preferred from an analytical point of view, as this produces more concentrated samples for analysis. Increased recovery is typically obtained either by using larger membrane surface area (longer microdialysis membrane) or by reducing perfusion flow-rates. Using longer membranes can however, compromise the spatial resolution (the anatomical sampling window), i.e., the sampling region dictates the length of the membrane. Using a too slow perfusion flow-rate may cause loss of temporal resolution due to the time needed to obtain sufficient sample for analysis. In some in vivo PK applications, a decrease in the number of time points may be unfavourable for drugs with relatively fast changes in their concentration. Thus, a suitable


balance between perfusion flow-rate and collection time interval has to be made in order to comply with the sensitivity of the analytical method regarding sample volume requirement and concentration detection limit (Davies, Cooper, Desmond, Lunte, & Lunte, 2000). In the current studies, a perfusion flowrate of 2 μl/min was found to provide an optimal compromise between recovery, sample volume and assay sensitivity, resulting in an acceptable temporal resolution for investigating the escitalopram and 5-HT concentration-time courses in the brain at a pharmacologically relevant dose. Besides the issue of dialysis recovery, the PK validation of the rat model revealed that brain extracellular concentrations of escitalopram were unaffected by catheterization of the carotid artery following systemic drug administration. Hence, a potential disruption of the integrity of the BBB with subsequent enhancement of the permeability into the brain, as observed following middle cerebral artery occlusion in rats (Belayev, Busto, Zhao, & Ginsberg, 1996), was not detected by microdialysis sampling of escitalopram. The additional effects of the surgical and experimental procedures on basal biochemical parameters showed that the surgical instrumentation associated with coupling of intracerebral microdialysis with automated blood sampling did not affect basal hippocampal 5-HT dialysate or plasma corticosterone levels. The observed absolute baseline concentrations of plasma corticosterone were in the same range as reported by others using conscious catheterized rats (Suzuki et al., 1997; Zuideveld, van Gestel, Peletier, Van der Graaf, & Danhof, 2002). In addition, the automated blood sampling itself did not evoke an increase in corticosterone levels during repetitive sampling, which was in agreement with previous investigations performed using this methodology in freely moving rats (Royo, Björk, Carlsson, Mayo, & Hau, 2004). Thus, the current findings indicate that the employment of automated blood sampling simultaneously with microdialysis is not associated with an increased stress response, and suggests that hippocampal 5-HT and plasma corticosterone levels should be viable measures of drug-induced effects in such integrated rat models. The evaluation of the automated blood sampling system for PK purposes showed that complete and reproducible concentration-time profiles could be obtained from plasma in individual animals. However, a loss in drug concentration of approximately 15% was evident following sample withdrawal by the automated sampling device. This might be attributed to a dilutional loss within the fluid lines and/or adsorption to tubings. As the drug loss was shown to be independent of the concentration tested, dilution appears to be the most likely cause of the partial recovery. As a consequence, applying such automated sampling systems for PK purposes should preferentially be accompanied by an assessment of drug loss for each compound to be studied. All together, key parameters that need to be considered when applying the current animal model for PK studies of new chemical entities includes appropriate microdialysis probe calibration, high analytical assay sensitivity and quantification of drug loss during automated sampling procedures. Also, when applying corticosterone as a measure of drug effect for PK/PD evaluation, sampling under non-stressful conditions is essential as the corticosterone levels were found to be susceptible to noise


C. Bundgaard et al. / Journal of Pharmacological and Toxicological Methods 55 (2007) 214–223

stimuli. Finally, a recovery period following surgical procedures of a minimum of two days was needed in order to obtain stable corticosterone levels (data not shown). The applicability of the model for multiple, simultaneous PK/PD investigations in individual animals was demonstrated after acute systemic administration of escitalopram. Hence, complete time-courses of escitalopram in brain and plasma were obtained simultaneously with the serotonergic and neuroendocrine pharmacological activity, measured as temporal changes in hippocampal 5-HT and plasma corticosterone levels. Compared to the PK of racemic citalopram (R(−) and S(+) citalopram) previously reported in rats (Cremers et al., 2000), the PK of escitalopram (the S(+)-enantiomer) exhibited a higher clearance and a shorter half-life in plasma. This might be explained by a more rapid demethylation of escitalopram as compared to R-citalopram as found both in vitro and in vivo (Kugelberg, Carlsson, Ahlner, & Bengtsson, 2003; Olesen & Linnet, 1999). In the brain, escitalopram reached maximal levels very rapidly. Thus, the highest brain extracellular concentration of escitalopram was recorded in the first dialysate sample, i.e., less than 10 min after intravenous drug administration. The parallel decline of unbound escitalopram in brain and plasma suggests a rapid exchange and equilibration between these two compartments. These observations were accompanied by an immediate effect on the brain extracellular 5-HT levels reaching approximately 400% relative to baseline levels consistent with earlier findings after s.c. administration (Ceglia et al., 2004; Mørk, Kreilgaard, & Sanchez, 2003). The functional response of the increased serotonergic activity following escitalopram administration on the HPA-axis was implicated by a short-lasting increase in corticosterone levels in plasma. This response pattern might reflect the activation of inhibitory feedback loops controlling the HPA-axis (Raap & Van de Kar, 1999). The temporal PK/PD correlations showed that hippocampal 5-HT output and corticosterone response did not correlate equally with the PK of escitalopram. Thus, combined with PK measures in plasma and brain in the vicinity of the pharmacological ‘site-of-action’, adaptive regulations of the serotonergic system following acute SSRI treatment can be investigated and related to drug exposure using the current animal model. Such a mechanistic PK/PD model has recently been developed for investigations of the acute serotonergic feedback effects following i.v. administration of escitalopram applying the integrated microdialysis rat model (Bundgaard, Larsen, Jørgensen, & Gabrielsson, in press). In conclusion, the applicability of the rat model was demonstrated for simultaneous characterization of multiple temporal relationships between escitalopram concentrations in plasma and brain extracellular fluid and neurochemical as well as neuroendocrine effects. As these data are obtained in conscious, freely moving animals without stress from sampling procedures or anaesthesia, PK/PD studies conducted this way are less subjected to experimental artifacts. In addition, as both PK and PD information are collected in the same individual animal, more reliable and accurate data would be expected in a reduced number of animals.

References Belayev, L., Busto, R., Zhao, W., & Ginsberg, M. D. (1996). Quantitative evaluation of blood–brain barrier permeability following middle cerebral artery occlusion in rats. Brain Research, 739, 88−96. Bouw, M. R., & Hammarlund-Udenaes, M. (1998). Methodological aspects of the use of a calibrator in in vivo microdialysis — further development of the retrodialysis method. Pharmaceutical Research, 15, 1673−1679. Bundgaard, C., Larsen, F., Jørgensen, M., & Gabrielsson, J. (in press). Mechanistic model of acute autoinhibitory feedback action after administration of SSRIs in rats: application to escitalopram-induced effects on brain serotonin levels. European Journal of Pharmaceutical Sciences. Burgio, D. E., & McNamara, P. J. (1993). Evaluation of recovery and loss from microdialysis probes of benzodiazepines and antipyrine in vitro. Pharmaceutical Research, 10(Suppl. 10), 349. Ceglia, I., Acconcia, S., Fracasso, C., Colovic, M., Caccia, S., & Invernizzi, R. W. (2004). Effects of chronic treatment with escitalopram or citalopram on extracellular 5-HT in the prefrontal cortex of rats: Role of 5-HT1A receptors. British Journal of Pharmacology, 142, 469−478. Cremers, T. I. F. H., de Boer, P., Liao, Y., Bosker, F. J., den Boer, J. A., Westerink, B. H. C., et al. (2000). Augmentation with a 5-HT1A, but not a 5-HT1B receptor antagonist critically depends on the dose of citalopram. European Journal of Pharmacology, 397, 63−74. Davies, M. I., Cooper, J. D., Desmond, S. S., Lunte, C. E., & Lunte, S. M. (2000). Analytical considerations for microdialysis sampling. Advanced Drug Delivery Reviews, 45, 169−188. de Lange, E. C., Ravenstijn, P. G., Groenendaal, D., & van Steeg, T. J. (2005). Toward the prediction of CNS drug-effect profiles in physiological and pathological conditions using microdialysis and mechanism-based pharmacokinetic–pharmacodynamic modeling. AAPS Journal, 7, E532−E543. Diehl, K. H., Hull, R., Morton, D., Pfister, R., Rabemampianina, Y., Smith, D., et al. (2001). A good practice guide to the administration of substances and removal of blood, including routes and volumes. Journal of Applied Toxicology, 21, 15−23. Drijfhout, W. J., Kemper, R. H., Meerlo, P., Koolhaas, J. M., Grol, C. J., & Westerink, B. H. (1995). A telemetry study on the chronic effects of microdialysis probe implantation on the activity pattern and temperature rhythm of the rat. Journal of Neuroscience Methods, 61, 191−196. Fagin, K. D., Shinsako, J., & Dallman, M. F. (1983). Effects of housing and chronic cannulation on plasma ACTH and corticosterone in the rat. American Journal of Physiology, 245, E515−E520. Fox, E., Bungay, P. M., Bacher, J., McCully, C. L., Dedrick, R. L., & Balis, F. M. (2002). Zidovudine concentration in brain extracellular fluid measured by microdialysis: steady-state and transient results in rhesus monkey. Journal of Pharmacology and Experimental Therapeutics, 301, 1003−1011. Groth, L., & Jørgensen, A. (1997). In vitro microdialysis of hydrophilic and lipophilic compounds. Analytica Chimica Acta, 355, 75−83. Gumbleton, M., & Benet, L. Z. (1991). Drug metabolism and laboratory anesthetic protocols in the rat: examination of antipyrine pharmacokinetics. Pharmaceutical Research, 8, 544−546. Gunaratna, P. C., Kissinger, P. T., Kissinger, C. B., & Gitzen, J. F. (2004). An automated blood sampler for simultaneous sampling of systemic blood and brain microdialysates for drug absorption, distribution, metabolism, and elimination studies. Journal of Pharmacological and Toxicological Methods, 49, 57−64. Kugelberg, F. C., Carlsson, B., Ahlner, J., & Bengtsson, F. (2003). Stereoselective single-dose kinetics of citalopram and its metabolites in rats. Chirality, 15, 622−629. Le Quellec, A., Dupin, S., Genissel, P., Saivin, S., Marchand, B., & Houin, G. (1995). Microdialysis probes calibration: gradient and tissue dependent changes in no net flux and reverse dialysis methods. Journal of Pharmacological and Toxicological Methods, 33, 11−16. Loch, J. M., Potter, J., & Bachmann, K. A. (1995). The influence of anesthetic agents on rat hepatic cytochromes P450 in vivo. Pharmacology, 50, 146−153. Lönnroth, P., Jansson, P. A., & Smith, U. (1987). A microdialysis method allowing characterization of intercellular water space in humans. American Journal of Physiology, 253, E228−E231.

C. Bundgaard et al. / Journal of Pharmacological and Toxicological Methods 55 (2007) 214–223 Mørk, A., Kreilgaard, M., & Sánchez, C. (2003). The R-enantiomer of citalopram counteracts escitalopram-induced increase in extracellular 5-HT in the frontal cortex of freely moving rats. Neuropharmacology, 45, 167−173. Olesen, O. V., & Linnet, K. (1999). Studies on the stereoselective metabolism of citalopram by human liver microsomes and cDNA-expressed cytochrome P450 enzymes. Pharmacology, 59, 298−309. Pardridge, W. M. (1998). CNS drug design based on principles of blood–brain barrier transport. Journal of Neurochemistry, 70, 1781−1792. Paxinos, G., & Watson, C. (1986). The rat brain in stereotaxic coordinates. San Diego: Academic Press. Plock, N., & Kloft, C. (2005). Microdialysis — theoretical background and recent implementation in applied life-sciences. European Journal of Pharmaceutical Sciences, 25, 1−24. Raap, D. K., & Van de Kar, L. D. (1999). Selective serotonin reuptake inhibitors and neuroendocrine function. Life Sciences, 65, 1217−1235. Royo, F., Björk, N., Carlsson, H. E., Mayo, S., & Hau, J. (2004). Impact of chronic catheterization and automated blood sampling (Accusampler) on serum corticosterone and fecal immunoreactive corticosterone metabolites and immunoglobulin A in male rats. Journal of Endocrinology, 180, 145−153. Sánchez, C., Bøgesø, K. P., Ebert, B., Reines, E. H., & Bræstrup, C. (2004). Escitalopram versus citalopram: the surprising role of the R-enantiomer. Psychopharmacology, 174, 163−176. Schuck, V. J. A., Rinas, I., & Derendorf, H. (2004). In vitro microdialysis sampling of docetaxel. Journal of Pharmaceutical and Biomedical Analysis, 36, 807−813. Shimizu, N., Take, S., Hori, T., & Oomura, Y. (1992). In vivo measurement of hypothalamic serotonin release by intracerebral microdialysis: significant enhancement by immobilization stress in rats. Brain Research Bulletin, 28, 727−734. Sidhu, J., Priskorn, M., Poulsen, M., Segonzac, A., Grollier, G., & Larsen, F. (1997). Steady-state pharmacokinetics of the enantiomers of citalopram and its metabolites in humans. Chirality, 9, 686−692. Song, Y., & Lunte, C. E. (1999). Comparison of calibration by delivery versus no net flux for quantitative in vivo microdialysis sampling. Analytica Chimica Acta, 379, 251−262. Stenken, J. A. (1999). Methods and issues in microdialysis calibration. Analytica Chimica Acta, 379, 337−358. Suzuki, K., Koizumi, N., Hirose, H., Hokao, R., Takemura, N., & Motoyoshi, S. (1997). Blood sampling technique for measurement of plasma arginine


vasopressin concentration in conscious and unrestrained rats. Laboratory Animal Science, 47, 190−193. Tegeder, I., Muth-Selbach, U., Lötsch, J., Rüsing, G., Oelkers, R., Brune, K., et al. (1999). Application of microdialysis for the determination of muscle and subcutaneous tissue concentrations after oral and topical ibuprofen administration. Clinical Pharmacology and Therapeutics, 65, 357−368. Thase, M. E. (2006). Managing depressive and anxiety disorders with escitalopram. Expert Opinion on Pharmacotherapy, 7, 429−440. Thorré, K., Pravda, M., Sarre, S., Ebinger, G., & Michotte, Y. (1997). New antioxidant mixture for long term stability of serotonin, dopamine and their metabolites in automated microbore liquid chromatography with dual electrochemical detection. Journal of Chromatography. B, Biomedical Sciences and Applications, 694, 297−303. Vallée, M., Mayo, W., Maccari, S., Le Moal, M., & Simon, H. (1996). Long-term effects of prenatal stress and handling on metabolic parameters: relationship to corticosterone secretion response. Brain Research, 712, 287−292. van Dongen, J. J., Remie, R., Rensema, J. W., & van Wunnik, G. H. J. (1990). Manual of microsurgery on the laboratory rat. Maastricht: Elsevier Science Publishers B.V. Wang, S. C., Ho, L. K., Yen, J. C., & Tsai, T. H. (2006). An automated blood sampling system to measure lovastatin level in plasma and faeces. Biomedical Chromatography, 20, 911−916. Watanabe, K., Matsuka, N., Okazaki, M., Hashimoto, Y., Araki, H., & Gomita, Y. (2002). The effect of immobilization stress on the pharmacokinetics of omeprazole in rats. Acta Medica Okayama, 56, 19−23. Wong, Y. N., Chien, B. M., & D'mello, A. P. (1994). Analysis of corticosterone in rat plasma by high-performance liquid chromatography. Journal of Chromatography. B, Biomedical Sciences and Applications, 661, 211−218. Woodward, C. J., & Emery, P. W. (1987). Determination of plasma corticosterone using high-performance liquid chromatography. Journal of Chromatography, 419, 280−284. Zuideveld, K. P., van Gestel, A., Peletier, L. A., Van der Graaf, P. H., & Danhof, M. (2002). Pharmacokineticpharmacodynamic modelling of the hypothermic and corticosterone effects of the 5-HT1A receptor agonist flesinoxan. European Journal of Pharmacology, 445, 43−54.