Determination of Atomoxetine or Escitalopram in human plasma by HPLC. Applications in Neuroscience Research Studies

Abstract

Background

Atomoxetine and escitalopram are potent and selective drugs approved for noradrenergic or serotonergic modulation of neuronal networks in attention-deficit hyperactivity disorder (ADHD) or depression, respectively. High-performance liquid chromatography (HPLC) methods still play an important role in the Therapeutic Drug Monitoring (TDM) of psychopharmacological drugs, and coupled with tandem mass spectrometry are the gold standard for the quantification of drugs in biological matrices, but not available everywhere. The aim of this work was to develop and validate a HPLC method for neuroscientific studies using atomoxetine or escitalopram as a test drug.

Methods

A HPLC method from routine TDM determination of atomoxetine or citalopram in plasma was adapted and validated for use in neuroscientific research. Using photo diode array detection with UV absorption at 205 nm, the variation of internal standard within one chromatographic method enables separate drug monitoring for concentration-controlled explorative studies in healthy humans and patients with Parkinson’s disease.

Results

The method described here was found to be linear in the range of 0.002 - 1.4 mg/L for atomoxetine and 0.0012 - 0.197 mg/L for escitalopram, with overall mean intra-day and inter-day imprecision and accuracy bias < 10% for both drugs. The method was successfully applied in concentration-controlled neuroimaging studies in populations of healthy humans and patients with Parkinson’s disease.

Conclusion

A simple, sensitive, robust HPLC method capable of monitoring escitalopram and atomoxetine is presented and validated, as a useful tool for drug monitoring and the study of pharmacokinetics in neuroscientific study applications.

Introduction

The study of pharmacokinetics in healthy humans as well as in patients aimed to demands of bioavailability, drug monitoring, and drug interactions or in special risk situations like restricted organ functions requires selective, sensitive, efficient and robust analytical techniques. Analytical methods established during clinical drug research programs of phase 1 to phase 3 are often directed to the specific parent drug and main metabolites and later adapted for requirements of clinical practice when there a medical need exists. Within the patient care, recommendations for measuring plasma drug concentrations for psychopharmacologic drugs especially are given, based on levels of evidence [1]. In contrast, within drug development only few clinical trials are designed in a randomized, concentration-controlled manner [2]. However, the implementation of a ‘concentration – effect’ relationship over a ‘dose – effect’ relationship may be even more important in clinical research, especially in neuroscience. In neuro-psychopharmacologic drug development, neuro-receptor ligand positron-emission tomography (PET) represents an established design for comparison of plasma vs brain kinetics. In contrast, pharmaco-fMRI studies are less invasive and hence very attractive if reliable reproducibility is succeeded. To achieve this, among numerous inevitable factors in either case a comprehensible and controlled drug monitoring is necessary, which is often not yet implemented. The exposure of healthy humans or patients to neuropsychopharmacological drugs for exploring test paradigms coupled with functional magnetic resonance imaging (fMRI) studies of the central nervous system is a common design in neuroscience research [3]. In this setting, the measurement and monitoring of test drug concentrations serves not only as a control of placebo or verum conditions in double-blinded, randomized trials. Rather, it also represents an essential co-variable in the correlational analysis between task-related performance results and the corresponding image presentation of increased neuronal activity in specific neuro-anatomic structures or functional clusters within e.g. the targeted serotonergic or noradrenergic nervous system.

High-performance chromatographic methods play an important role in the determination of psychopharmacological drugs, but commercial assays are limited in their number and the selection of drugs for which assays are available. On the other hand, quality-assured chromatographic routine methods for use in therapeutic drug monitoring (TDM) may be adopted, to and thus applied for, particular scientific purposes.

Aiming at a highly specific pharmacological modulation on neurotransmitter levels of neuronal networks involved in cognition, learning, memory, impulsivity, behavior, and functional execution, the most selective and potent Specific Serotonin-Reuptake-Inhibitor (SSRI), escitalopram, and Specific Noradrenergic Reuptake-Inhibitor (SNRI), atomoxetine, are nowadays often preferred over other approved drugs. Since 2004, atomoxetine is approved in Europe for attention-deficit hyperactivity disorder (ADHD) in children of age over 6 years and adolescents, and since 2013 also in adult patients within a comprehensive psycho-social treatment program. The recommended oral dosage can be titrated from 10 to 80 mg once and spread into two doses daily. In healthy volunteers, atomoxetine has an absolute oral bioavailability of 63-94% which is unaffected by food. It is highly bound to albumin (98%) and extensively hydroxylated and N-des-methylated, predominantly by CYP2D6 and to a minor extent by CYP2C19, in the liver. The active 4-hydroxy-atomoxetine metabolite is rapidly processed to its respective glucuronide, so that the circulating amount of the active metabolite is minimal, and after multiple doses of atomoxetine the exposure of the parent drug was approximately 35 times larger than that of the metabolite [4]. Maximum plasma concentrations are reached between 1-2 hours after intake of 1.2 mg/kg body weight per day. The mean half-life is 4 hours while CYP2D6 poor metabolizers show a prolongation up to 21 hours. The therapeutic reference range is given at 200-1000 ng/ml [1].

Escitalopram, introduced in Europe in 2002/2003, is the active (S)-enantiomer of the SSRI citalopram and one of the most frequently used first-line drugs for major depression disorder. Other recognized indications are panic disorders, generalized or social phobia, and obsessive-compulsive disorders. Besides this, already a single dose of 20 mg escitalopram has been found in healthy subjects to dramatically alter functional connectivity throughout the whole brain, suggesting a key role for the serotonin transporter in the modulation of the functional macroscale connectome [5]. The pharmacokinetics of escitalopram has been extensively reviewed [6]. Escitalopram has a bioavailability of 80% and is widely distributed into various tissues, with an apparent distribution volume of ~1100 L in the terminal phase after oral administration. Metabolic clearance is affected by CYP2C19 and to a lesser extent by CYP2D6, CYP3A4, age and gender. N-des-methylated metabolites may contribute only weakly to its pharmacological action. The half-life varies between 27-32 hours. Its therapeutic reference range, which indicates C_max values measured in remitters from depression, was determined at 15-80 ng/ml [1], while more recent data give some evidence for a putative therapeutic threshold of 20 ng/ml [7]. Escitalopram concentrations were found to be significantly higher in patients treated with the known inhibitors of CYP2C19, omeprazole and esomeprazole [8]. According to the Consensus Guidelines for Therapeutic Drug Monitoring in Neuro-psychopharmacology [1], a drug monitoring of escitalopram is recommended for dose titration and for special indications or problem solving (level 2). For instance, QT interval prolongation associated critical incidents and again stronger clinical ECG controls in initialization of escitalopram therapy fortified the role of TDM [9]. The recommendation for atomoxetine is at level 3 in terms of being useful in special indications or problem solving.

Several mass spectrometry- or high performance liquid chromatography (HPLC)-based methods have been published for atomoxetine or escitalopram, neither of those has been specified as suitable for both drugs. Though, escitalopram in plasma has been determined by several LC-MS methods, such as LC-ESI-MS with LLOQ 1 ng/ml following L/L extraction or directly electro-sprayed into a (time of flight) TOF-MS [10,11] or with LLOQ of 0.2 ng/ml by LC-Tandem MS following liquid extraction and evaporation [12,13], or liquid extraction plus SPE clean up and gradient flow conditions [14] as well as direct injection following protein denaturation and evaporation [15], both without data on method evaluation. More toxicological and forensic directed LC-MS methods for determination of escitalopram in neonatal hair (LOQ 25 pg/ml) [16] or in postmortem human samples [17] have also been reported. Focusing on sensitive but simple HPLC methods for escitalopram in human plasma avoiding fluorescence detection with derivatization procedure [18,19] with LOD of 0.2 to 6 ng/ml after SPE extraction, only one HPLC/UV (210 nm) method [20] by using L/L extraction plus evaporation reported a LOD of 1 ng/ml and a lowest calibration point of 2.5 ng/ml, achieved only by a large injection volume of 400µl chromatographed on a C18 reversed phase column (4.6 x 150 mm).

Similarly, while one published LC-MS method for determination of atomoxetine in human plasma reported to be able to quantify a minimum of 0.25 ng/ml bearing data on method validation and performance [4], other LC-Tandem MS methods achieved a LOD of 0.5 ng/ml atomoxetine [21, 22] using liquid extraction whereby could also identify and quantify comedications taken by the patients recruited in the study like carbamazepine and trans-10,11-dihydro-CBZ, sertraline and desmethylsertraline, risperidone and 9-hydroxyrisperidone. However, the literature reference for method validation given by authors was not adequate. Again, focusing on sensitive HPLC methods, avoiding fluorescence detection following derivatization [23,24] only two published methods provided information’s regarding LLOD of 2.5 ng/ml [25] and 50 ng/ml [26], both using liquid-liquid extraction and evaporation under nitrogen and separation on a 4.6 x 150 mm (i.d.) C18 reversed phase chromatographic column with detection at 210 or 272 nm, respectively. Simultaneous detection of atomoxetine and escitalopram by HPLC methods has not been published so far. Although, mass spectrometry in principal capable to achieve this, it is expensive, requires specially qualified personnel and is often not available in smaller laboratories, including clinical research institutes. Recognizing this data availability, we therefore consequently developed a simple but very sensitive isocratic HPLC method with UV detection capable of determination either escitalopram or atomoxetine in human plasma that is alternating applicable by a simple switch from one internal standard to the second internal standard. To attain an acceptable short run time, low flow rate and high absorption signals, we preferred a narrow bore (2.1 x 150 mm) C8 reversed phase chromatographic column and a reliable selective liquid/liquid extraction for preparation of small 0.5 ml plasma sample without adding cost raising solid phase extraction. Hence, we present the detailed description and extensive validation of an HPLC method for the analysis of escitalopram or atomoxetine in human plasma. The method has been adapted (downsized) from routine TDM to the specific requirements of clinical neuroscience studies, and successfully proved in clinical research projects in healthy humans and patients with mild to moderate Parkinson’s disease.

Methods

Chemicals and Materials

Atomoxetine hydrochloride, desipramine hydrochloride, escitalopram oxalate, mianserine hydrochloride, n-hexane and dimethyl sulfoxide (DMSO) were purchased from Sigma Aldrich Chemie (Steinheim, Germany). Water HiPerSolv CHROMANORM for HPLC and acetonitrile HiPerSolv CHROMANORM for HPLC Gradient Grade, iso-amyl alcohol, potassium dihydrogen phosphate ultrapure and hydrochloric acid (HCl) 36.5-38.0% were purchased from VWR International (Darmstadt, Germany), and sodium hydroxide (NaOH) for analysis was from Merck (Darmstadt, Germany). Blank human blood was collected from healthy, drug-free voluntary blood donors and checked on bio-safety by the university hospital department of Transfusion Medicine. Plasma was obtained by centrifugation of blood treated with sodium heparin as anticoagulant. Pooled plasma was prepared and stored at -80°C until needed.

Instrumentation

Analyses were performed using a modular Agilent 1100 gradient system (Agilent Technologies, Palo Alto, CA, USA), consisting of degasser, column heater, photo diode array detector and ChemStation software for data acquisition and analysis.

Chromatographic Conditions

Chromatographic separation was performed on a Zorbax Eclipse XDB C8, 3.5-Micron Narrow-Bore column (2.1 x 150 mm) equipped with a Narrow-Bore guard column 2.1 x 12.5 mm 5-Micron (Agilent Technologies) and maintained at 25°C. Acetonitrile and aqueous 30 mM potassium dihydrogen phosphate (34:66 (v/v), pH 5.1) were used as the mobile phase, at a flow rate of 0.225 mL/min., resulting in a column back pressure at about 123 bars. The wavelength of the detector was set to 205 nm. Chromatographic capacity factor (k’) for the early eluting analyte escitalopam was calculated with 2.7 which is sufficient to demonstrate, that chromatography has efficiently taken place.

Preparation of stock solutions, calibrators and quality control samples

Both escitalopram and atomoxetine were precisely weighed and dissolved in DMSO to obtain stock solutions with a concentration of 2 mg/mL of each analyte, after correction for the purity factor.

To obtain working solutions, defined volumes of the respective stock solution were pipetted to volumetric flasks and diluted with water. Before dilution, the net mass of each pipetted volume was divided by the density to calculate the true volume. Exemplarily, the dilution schedule for atomoxetine working solutions is depicted in Table 1. Target volumes and the resulting target concentrations are shown. Real values obtained by calculation using density and mass differ more or less from the nominal values.

Table 1. Preparation of atomoxetine working solutions.

Calibration Working Solution	Source Solution	Concentration of Source Solution (μg/mL)	Volume of Source Solution	Final Volume of Working Solution (mL)	Final Concentration of Working Solution (μg/mL)
WS1	Stock Solution	2001.8	146.75	2	146.887
WS2	Stock Solution	2001.8	10.02	5	4.011
WS3	WS2	4.01	199.97	2	0.401

The calibration samples were prepared as bulk solutions by spiking drug-free human plasma with a working solution of appropriate concentration to obtain nominal concentrations in the range of 1 - 200 ng/mL escitalopram (1.05, 2.11, 3.14, 9.70, 48.28, 96.56, and 196.76 ng/mL) and 2 - 1500 ng/mL atomoxetine (1.97, 3.93, 7.86, 15.73, 39.32, 117.97, 576.05, and 1440.12 ng/mL), respectively according to the recommendation of the EMA guideline on bioanalytical method validation covering a calibration curve range from LLOQ to ULOQ. Exemplarily, the preparation of atomoxetine calibration samples is given in Table 2. Volumes of the source solutions were added to pre-labeled polypropylene tubes. The solutions were diluted by adding drug-free plasma using a fixed-volume pipette. The solutions were mixed well using a vortex mixer. 0.5 mL portions of each pooled standard were aliquoted into pre-labeled 2 mL polypropylene vials. All calibration samples were stored in a freezer set to maintain -20°C. Quality control samples were prepared in the same manner, using separate stock solutions, to yield nominal concentrations of 1, 4, 25, and 50 ng/mL escitalopram and 2, 4, 40, and 400 ng/mL atomoxetine. Internal standard solutions were prepared in water at final concentrations of 2.0 µg/mL desipramine and 1.9 µg/ml mianserine, respectively, by dilution of aqueous stock solutions.

Table 2. Preparation of atomoxetine calibration samples (C0 – C7 target concentrations).

Calibration Sample ID	Source Solution ID	Conc. of Source Solution (μg/mL)	Volume of Source Solution (μL)	Volume of Drug-Free Plasma (mL)	Calibration Samples Concentration (ng/mL)
CO	WS3	0.401	12.5	2.5375	1.97
C1	WS3	0.401	25	2.525	3.93
C2	WS2	4.011	5	2.545	7.86
C3	WS2	4.011	10	2.54	15.73
C4	WS2	4.011	25	2.525	39.32
C5	WS2	4.011	75	2.475	117.97
C6	WS1	146.887	10	2.54	576.05
C7	WS1	146.887	25	2.525	1440.12

Sample Preparation

Various sample preparation procedures were tested, including single liquid extraction with consecutive evaporation under Nitrogen and reconstitution and SPE as well. Due to similar pKa values of the analytes (atomoxetine and escitalopram 9.8, mianserine 6.92, desipramine 10.0), high pH liquid extraction into n-hexane (logP values of all analytes range between 3.5 and 4.0) and subsequent low pH re-extraction into 0.01 M HCl was more selective and sensitive than all other tested, achieved acceptable percentage recovery and is not expensive. 0.5 mL plasma, 20 µL internal standard, 20 µL iso-amyl alcohol and 1.5 mL n-hexane were vortexed for 5 minutes. After centrifugation for 3 minutes at 4,000 x g, the supernatant was transferred to a 2 mL vial and re-extracted by adding 0.15 mL 10 mM aqueous HCl and vigorous shaking for 5 minutes. After centrifugation at 12,500 x g for 3 min, the lower aqueous phase was collected and 10 µl were injected into the HPLC system.

Results

Applying the method described above on plasma samples from healthy volunteers and patients treated with either escitalopram or atomoxetine, characteristic chromatograms were obtained as shown in Figure 1. Retention times (t_R) for escitalopram and the corresponding internal standard desipramine were found to be 5.8 and 8.1 minutes, respectively, while the drug peaks for atomoxetine and its corresponding internal standard mianserine were found at 8.8 minutes and 6.5 minutes, respectively.

Figure 1.

Overlay of chromatograms of typical in vivo human plasma samples: (A) blank plasma (pre-dose), (B) 47.1 ng/mL Escitalopram (1), ISTD Desipramine (2) and (C) 520.9 ng/mL Atomoxetine (4), ISTD Mianserine (3).

Validation

Validation was performed according to the EMA Guideline on bioanalytical method validation [27]. Each calibration curve (Figures 2–3) was generated using human plasma spiked with 7 concentrations of each analyte, ranging from 1.21 - 196.76 ng/mL for escitalopram and 2.0 - 1440.12 ng/mL for atomoxetine, respectively. The linearity of the calibration curve range was evaluated by using a least-squares linear regression with 1/response weighting. The resulting calibration functions for escitalopram (y=0.01225473x-0.0053631) and atomoxetine (y=0.01614966x-0.0233627) showed corresponding correlation coefficient (r) of 0.99989 for escitalopram and 0.99961 for atomoxetine. Calibration statistic data are given in Table 3. The lowest calibrator of 1.22 ng/mL escitalopram and 2.0 ng/mL atomoxetine, respectively, represented the lower limit of quantification (LLOQ). Results on within-run and between-run precision and accuracy at LLOQ and the quality control samples LQC, MQC1, MOC2 and HQC are given in Table 4. The overall mean imprecision and bias were <10% for both escitalopram and atomoxetine. For specificity determination, blank plasma samples from different origin without internal standard were prepared as described above and evaluated for any interference with the retention times of the specific peaks of drugs and internal standards. The assay has been successfully applied to hemolized and lipemic samples without any impact on performance.

Figure 2.

Escitalopram calibration curve graphics

Figure 3.

Atomoxetine calibration curve graphics

Table 3. Escitalopram and atomoxetine calibration data statistics.

Analyte	Level	Concentration, theoretic ng/ml	Peak Area	Concentration, found ng/ml	Rel. Deviation %
Escitalopram	1	1.055	0.68812	1.00	-5.586
	2	2.110	1.92400	2.03	-3.636
	3	3.140	3.13469	3.07	-2.243
	4	9.700	10.87884	9.66	-0.459
	5	48.28	60.67636	47.97	-0.636
	6	96.56	125.37290	97.56	1.033
	7	196.76	210.89548	196.32	-0.224
ISTD (Desipramine)	1	1.0	100.55300
	2	1.0	98.39336
	3	1.0	97.18863
	4	1.0	96.30493
	5	1.0	104.15946
	6	1.0	105.33930
	7	1.0	87.85548
Atomoxetine	1	1.97	0.591826	1.97	0.174
	2	3.93	2.658830	3.93	-0.119
	3	7.86	6.620210	7.70	-1.985
	4	15.73	13.776890	14.52	-7.691
	5	39.32	36.137170	36.73	-6.586
	6	117.97	123.014380	113.03	-4.188
	7	576.05	650.324580	594.31	3.170
	8	1440.12	1601.68542	1430.76	-0.650
ISTD (Mianserine)	1	1.0	69.56408
	2	1.0	66.42089
	3	1.0	65.51193
	4	1.0	65.25198
	5	1.0	63.41822
	6	1.0	68.26469
	7	1.0	67.92218
	8	1.0	69.38823

Table 4. Results of Within- and Between-Run Escitalopram and Atomoxetine Quality Controls.

	LLOQ	LQC	MQC1	MQC2	HQC
Escitalopram Nominal Concentration (ng/mL)	1.22 ng/mL	3.63 ng/mL	25.18 ng/mL	66.22 ng/mL	132.83 ng/mL
Within-run
Imprecision (CV%)	7.3	4.4	5.8	4.3	5.8
Accuracy (%)	106.5	101.0	97.2	99.0	97.8
Between-run
Imprecision (CV%)	8.2	2.2	3.7	5.2	4.5
Accuracy (%)	102.2	100.3	100.5	100.4	100.6
Atomoxetin Nominal Concentration (ng/mL)	2.0 ng/mL	3.69 ng/mL	36.91 ng/mL	413.78 ng/mL	873.22 ng/mL
Within-run
Imprecision (CV%)	3.6	3.4	5.9	5.6	2.3
Accuracy (%)	101.4	99.6	94.7	99.3	97.4
Between-run
Imprecision (CV%)	1.1	2.1	0.7	1.9	1.9
Accuracy (%)	100.8	98.8	97.2	99.0	99.4

LLOQ, lower limit of quantification of the method; LQC, low quality control; MQC, median quality control; HQC, high quality control; CV, coefficient of variation

For the two internal standard peaks, the ratio of baseline interference peaks versus the mean peak area of six escitalopram or atomoxetine LLOQ samples was less than 0.20, and less than 0.05 in 24 different blank and zero samples tested. Selectivity was not systematically examined for the presence of interfering drugs with similar UV absorbance such as tri- or tetracyclic antidepressants including the internal standards used in this assay. However, in drug naïve volunteers and patients with Parkinson’s disease primarily treated with dopaminergic medication, interfering peaks were not expected and were actually not detected in any sample.

The mean absolute recovery of escitalopram, as determined by comparing the measured peak area of QC samples (n=6 each) with those of the extracted blank plasma samples spiked with reference standard at the same concentration, was found to be 97.2 ± 7.0, 93.5 ± 6.2, 97.7 ± 5.9, 99.5 ± 3.1 and 98.3 ± 2.6%, at 1.22 (LLOQ), 3.62 (LQC), 25.18 (MQC1), 66.22 (MQC2), and 132.83 ng/mL (HQC), respectively. For atomoxetine, values were determined at 79.7 ± 3.9, 84.9 ± 1.9, 82.0 ± 4.9, 80.3 ± 2.1 and 83.4 ± 3.7% at 2.0 (LLOQ), 3.69 (LQC), 36.91 (MQC1) and 413.78 (MQC2), and 873.22 ng/mL (HQC), respectively. Recoveries of desipramine (internal standard for escitalopram) and mianserine (internal standard for atomoxetine) were 94.1 ± 5.0% and 90.3 ± 1.0%, respectively. After injection of the highest calibrators (196.76 for escitalopram and 1440.12 ng/mL for atomoxetine), no carry-over in a blank sample was detected. Sample re-analysis showed acceptable long-term stability of 7 months in human plasma stored at -20°, as indicated by a decrease in concentration of < 5%. Stock solutions and working solutions of the analytes escitalopram and atomoxetine were confirmed to be stable in the refrigerator (4 - 8 °C) for 12 and 3 months, respectively.

Application

All human healthy volunteers and patients gave written informed consent to take part in various neuroimaging studies, approved by institutional ethical review boards [5, 28, 30–36]. The method described here for target drug and corresponding internal standard quantitation was applied to the determination of escitalopram in different study groups of healthy volunteers (mean age: 25, SD 7.2 years), comprising a total number of 125 individuals [5, 28–30]. Following single oral administration of 20 mg, plasma samples obtained at different time points (2.0, 2.5, 3.0, 3.5 and 5.5 hours post dosing) were determined as well as after repeated dosing for reaching the steady state on day 7 [30].

Mean escitalopram plasma levels (Fig. 4A) following oral single dose of 20 mg in a rather homogeneous population of young healthy human volunteers [5, 29–30] showed only minor time-related differences between sample time points at 2.0, 2.5, and 3.0 hours post dose. At 3.5 hours, a maximum plasma level was observed (mean value 23.6 ng/mL +/- 9.0), which was still well below the steady state plasma levels (44.1 ng/mL) on day 7 (168 hours). The obtained time course of escitalopram plasma concentrations in healthy volunteers from different neuroscience studies largely reflect and confirm published data on the drug absorption phase to reach the C_max in man [6], as also documented in the summary of product characteristics [38].

Figure 4.

Time course of geometric mean plasma levels of Escitalopram (A) upon a single oral dose application of 20 mg in healthy humans at various anticipated Tmax in separate studies (each symbol represent a distinct study), or following repeated dose application at steady state. (B) Atomoxetine plasma concentration time course (open symbols) with standard deviations and medians (closed symbols), following a single dose application of 40 mg in healthy humans (circles), patients with Parkinson’s disease (quadrates), and (C) cocaine-dependent individuals (triangles). Data obtained and composed from several separate neuroscience studies (for details, see text).

Atomoxetine was determined upon a single oral dose application of 40 mg in healthy volunteers [31, 33], cocaine dependent individuals (CDI) as volunteers [33], and in patients with mild to moderate Parkinson’s disease [34–37].

Overall, 65 healthy humans (mean age: 27 years, SD 7.3), 27 CDI (mean age: 41 years, SD 7.4) and 128 patients with Parkinson’s disease (mean age: 66.3 years, SD 6.8) were exposed to 40 mg atomoxetine and plasma levels were studied.

For example, in a within-subject, double blind, placebo-controlled study design [31], atomoxetine showed improved inhibitory control and increased activation in the right inferior frontal gyrus when healthy volunteers attempted to inhibit their responses (irrespective of success). Monitoring atomoxetine plasma level following single oral dose administration at an estimated T_maxbetween 1.5 and 3.0 hours post dosage in these 19 young human volunteers evidenced a significant correlation between plasma levels of atomoxetine and right inferior frontal gyrus activation during successful inhibition only. Several further studies on the involvement of noradrenergic mechanisms in attention, cognition and behavioral deficits used the same single dose of atomoxetine as a pharmacological challenge to examine the change in performance effects in neuro-psychological test batteries as a function of atomoxetine plasma levels. A total of 57 healthy volunteers [32, 33] and 128 patients were analyzed [34–37].

As shown in Figure 4 B, mean atomoxetine plasma levels in human volunteers were associated with extreme inter-individual variabilities, with some standard deviations being almost in the magnitude of the mean values. While the concentration time course of mean atomoxetine plasma level was highest already at 1.0 hour post dose, the median values point to the fact that in a substantial number of healthy humans the T_max may be reached only between 1.5 and 2.5 hours. This large variability in atomoxetine plasma concentrations also applies to older patients with Parkinson’s disease. In this population, however, the mean and median atomoxetine plasma levels at 1.0 hour are slightly lower and increase until 2.0 hours, reaching a higher mean level compared to the healthy humans which is still observed at 3 hours.

A comparable variation in atomoxetine plasma levels was found in the CDI volunteers (Figure 4 C); however, divergent low mean and median levels at 1 hour and very high peak levels at 2.5 hour post dose were observed.

Discussion

This paper describes the adaption and validation of a routine isocratic HPLC method with UV-based detection, for the measurement of atomoxetine or escitalopram by using different internal standards. This facile method described here, only requiring standard HPLC equipment, was specifically adapted for use in neuroscience research. It allows for concentration-controlled drug monitoring study design using the most relevant and highly specific key serotonergic and noradrenergic reuptake inhibitors in human research.

The method was found linear within the range of 1.22 - 196.7 ng/mL (atomoxetine) and 2.0 -1440 ng/ml (escitalopram), which is sufficient for single dose administration, and also allows for the description of the complete plasma kinetics in healthy volunteers and patients. The lower limit of quantification achieved for escitalopram and atomoxetine is nearly as low as in published LC-MS/MS methods [4, 12–13, 21]. This is particularly noteworthy since the mass spectrometry-based analysis of drugs requires expensive technical equipment and highly skilled personnel. This may represent a high barrier for its implementation in neuroscientific research institutions, due to the need of involving specialized external TDM laboratories. In contrast, the presented HPLC method offers the option for comparable facile and less laborious in-house analysis of atomoxetine and escitalopram.

Over several years, the method described here has been used in several studies for analyzing escitalopram in healthy volunteers, in which monitoring of escitalopram levels was a planned part of the study design to ensure compliance during placebo-controlled, double-blinded study conditions. More importantly, it allowed for exploring to what extent and under which conditions beneficial or detrimental neuro-physiological concentration-related effects of the drug were associated with the respective changes of neural BOLD (Blood-oxygen-leveldependent imaging) signal.

For example, this is particularly relevant in the treatment of patients with complex pathophysiological conditions like Parkinson’s disease when addressing the question of how behavioral dysfunctions involved in the disease are concentration-dependently affected by drugs like atomoxetine as a noradrenergic enhancer (promotor) on the neurotransmitter system level. Despite anticipated gastrointestinal dysfunctions like dysphagia, gastroparesis and constipation, which are commonly associated with PD due to autonomic deficits and lead to prolonged transit time and malabsorption in the gut [39], a delayed T_max (2 hours) compared to younger healthy humans was not observed. This may be caused by therapeutically substituted dopaminergic drugs like levodopa/carbidopa, entacapone, pramipexole, ropinirole, rotigotine corresponding to levodopa equivalent daily doses between 630 to 1080 mg/day.

The prolonged time concentration course of mean plasma levels between 2 - 6 hours may be explained by a moderately reduced age-related hepatic clearance and/or diminished body weight [40] in this older age, even though no direct evidence was obtained from these studies. According to Sauer et al. (2005), maximum plasma concentrations of atomoxetine in healthy humans may be estimated between 1.0 and 2.0 hours after oral application. This is confirmed by the observed median values, while mean values show extreme inter-individual variability. The fact that in a substantial number of healthy humans atomoxetine seems to be more rapidly absorbed may therefore again be of special interest in planning the design of concentration-controlled neuroscience paradigms. In CDI volunteers, no apparent explanation can be given for the remarkably low mean atomoxetine level at 1.5 hours, followed by very high levels at 2.5 hours post dose, when compared to non-cocaine dependent healthy volunteers or patients with Parkinson’s disease. As absolute oral bioavailability of atomoxetine ranges from 63 to 94%, which is governed by the extent of first pass metabolism, somatic features in drug disposition, hemodynamics, and clearance processes of atomoxetine caused by ongoing longtime cocaine use (confirmed by positive test results in urine samples) may be causally involved. Additionally, life style related differences in food intake may also have contributed, since food, albeit not affecting the extent of absorption, decreases C_max and delay times to reach maximum plasma concentration by 3 hours [40].

Conclusion

A simple, sensitive, robust HPLC method capable of escitalopram and atomoxetine monitoring in human plasma is presented, particularly useful and widely proven e.g. in neuroscience research as exemplified by several pharmacological neuroimaging studies in human volunteers as well as patients with Parkinson’s disease. Data on atomoxetine highlight the great inter-individual variation in plasma concentration after single dose application in human volunteers and patients with Parkinson’s disease with implications for a rationale study design.

Statement on Effects of Ethnicity.

This analytical in vitro method described here in its validation performance is not affected by ethnicity. Quite in contrast the method is suitable to map the different pharmacokinetics in different ethnicities based on variations in genotype of metabolizing cytochromes CYP2D6 which or CYP2C19 relevant for atomoxetine and escitalopram. In this study, plasma concentrations from healthy volunteers, cocaine-dependent individuals and patients with Parkinson’s disease were predominantly obtained from a Caucasian population.