S

A

SD

a

ARR2AA

KTOPLV

1

cTMcccmr

owlm1ptv

sa

T7

0d

Journal of Pharmaceutical and Biomedical Analysis 58 (2012) 177–181

Contents lists available at SciVerse ScienceDirect

Journal of Pharmaceutical and Biomedical Analysis

journa l homepage: www.e lsev ier .com/ locate / jpba

hort communication

n LC–MS/MS method for the determination of ofloxacin in 20 �l human plasma

andra A. Meredith ∗, Peter J. Smith, Jennifer Norman, Lubbe Wiesnerivision of Clinical Pharmacology, University of Cape Town, Medical School, Observatory, 7925 Cape Town, South Africa

r t i c l e i n f o

rticle history:eceived 11 April 2011eceived in revised form3 September 2011ccepted 26 September 2011vailable online 1 October 2011

a b s t r a c t

A sensitive and selective liquid chromatography–tandem mass spectrometry method was developedand validated for the determination of ofloxacin in 20 �l human plasma over the concentration rangeof 0.078–20 �g/ml. Sample preparation was achieved by protein precipitation with acetonitrile andmethanol containing the internal standard (Gatifloxacin). Chromatographic separation was achieved ona Luna 5 �m PFP (110 A, 50 × 2 mm) column with acetonitrile and water containing 0.1% formic acid(50:50, v/v) as the mobile phase, at a flow rate of 400 �l/ml. The within-day and between-day precision

eywords:uberculosisfloxacinrotein precipitationC–MS/MSalidation

determinations for ofloxacin, expressed as the percentage coefficient of variation, were lower than 7% atall test concentrations. Recovery of ofloxacin was greater than 70% and reproducible at the low, mediumand high end of the dynamic range. No significant matrix effects were observed for the analyte or internalstandard. The assay was successfully used to examine the pharmacokinetics of ofloxacin as part of a studyto characterize the pharmacokinetics of a number of anti-tuberculosis drugs utilized in the treatment ofmulti-drug resistant tuberculosis (MDR-TB).

. Introduction

The occurrence of multidrug-resistance (MDR) strains of tuber-ulosis (TB) contributes significantly to the increased incidence ofB. Ofloxacin is used as a second-line drug for the treatment ofDR-TB [1,2]. It is a broad-spectrum, chemotherapeutic antimi-

robial agent belonging to the Fluoroquinolone (quinolones) druglass. Quinolones target bacterial DNA gyrase, an enzyme thatontrols the supercoiling of DNA and separating of linked DNAolecules. Inhibition of DNA gyrase by the quinolones effects DNA

eplication, transcription, repair and recombination [3,4].A number of papers on the determination and quantification

f ofloxacin by High Performance Liquid Chromatography (HPLC)ith UV or fluorescence detection have been published, with the

imit of detection being around 20–25 ng/ml. The drawback of theseethods however, is that large plasma volumes (between 0.2 and

.0 ml) are required [3,5–7]. For pediatric patients this is not ideal,articularly when numerous samples are needed [8]. To minimisehe volume of blood sampling, drug assays requiring smaller sampleolumes are essential.

Advances in HPLC coupled with triple quadrupole masspectrometry (LC–MS/MS) technology have revolutionized drugnalysis in biological samples. Today LC–MS/MS is heralded the

∗ Corresponding author at: Division of Clinical Pharmacology, University of Capeown, Medical School, H50, Old Main Building, Groote Schuur Hospital, Observatory,925 Cape Town, South Africa. Tel.: +27 21 406 6479; fax: +27 21 406 6152.

E-mail address: sandra.meredith@uct.ac.za (S.A. Meredith).

731-7085/$ – see front matter © 2011 Elsevier B.V. All rights reserved.oi:10.1016/j.jpba.2011.09.030

© 2011 Elsevier B.V. All rights reserved.

gold standard for drug analysis, with vast improvements in sen-sitivity, selectivity and throughput [9]. The aim of this work was todevelop a sensitive LC–MS/MS method using low plasma volumesto facilitate pharmacokinetic studies and therapeutic drug moni-toring in paediatric patients. We describe here a robust LC–MS/MSmethod coupled with a simple protein precipitation extractionmethod for the determination of ofloxacin in 20 �l human plasma.

2. Experimental

2.1. Materials and chemicals

Methanol (LiChrosolv®), acetonitrile (LiChrosolv®), formic acid(pro analysis) and water (LiChrosolv®) were purchased fromMerck kGaA, Darmstadt, Germany. The ofloxacin reference stan-dard (EP) was supplied by Industrial Analytical (Kyalami, RSA)and gatifloxacin (internal standard) supplied by Toronto ResearchChemicals (TRC, Ontario, Canada). Drug free plasma was obtainedfrom Western Province Blood Transfusion Services at the GrooteSchuur Hospital, South Africa. A Phenomenex Luna, PFP(2) 100 A,50 mm × 2.0 mm column (Phenomenex, USA) was used for retain-ing ofloxacin and the internal standard.

2.2. Chemical structure

Chemical structure of ofloxacin is presented in Fig. 1.

178 S.A. Meredith et al. / Journal of Pharmaceutical an

2

bPfS

2

ioo((t0f0da

2

pimntfap2wpp

2

mwhaoct(

amm

Fig. 1. Chemical structure of ofloxacin.

.3. Instrumentation

The mobile phase was delivered with an Agilent 1200 seriesinary pump and the samples injected with an Agilent 1200 Higherformance Autosampler (Agilent, CA, USA). Detection was per-ormed by an AB Sciex API 3200 Q Trap mass spectrometer (ABciex, Ontario, Canada) fitted with a Turbo VTM ion source.

.4. Preparation of calibration

Calibration standards and quality controls (QC) were preparedn blank human plasma, and the preparation procedure performedn ice. Ofloxacin stock solution (SS1) was prepared in a mixturef acetonitrile and water (50:50, v/v) at 1000 �g/ml. Blank plasma10 ml) was spiked with SS1 for the highest calibration standard20 �g/ml) and serial dilutions were performed with blank plasmao attain the additional calibration standards (10, 5, 2.5, 1.25, 0.623,.313, 0.156 and 0.078 �g/ml). The same methodology was usedor the preparation of the high, medium and low QCs (16, 8 and.25 �g/ml, respectively). The calibration standards and QC stan-ards were briefly vortexed, aliquoted into polypropylene tubesnd stored at approximately −70 ◦C.

.5. Extraction procedure

The extraction procedure was performed on ice and inolypropylene test tubes. The plasma samples were thawed on

ce, briefly vortexed, and 20 �l placed into polypropylene tubes. Aixture of acetonitrile and methanol (80:20, v/v) containing inter-

al standard (gatifloxacin) at 200 ng/ml was added to precipitatehe proteins. The samples were vortexed for 1 min, ultrasonicatedor 5 min, vortexed for a further 30 s, and centrifuged for 5 mint 16,000 rcf. The supernatant (160 �l) was transferred to cleanolypropylene tubes and evaporated under nitrogen at 30 ◦C for0 min.. Mobile phase (100 �l), which consisted of acetonitrile andater with 0.1% formic acid (50:50, v/v), was added to the dry sam-les. The samples were vortexed for 30 s, transferred to a 96 wellolypropylene plate, and 5 �l were injected onto the HPLC column.

.6. Mass spectrometry

Electrospray ionization (ESI) was performed in the positive ionode with nitrogen as the nebulizing, turbo spray and curtain gasith the optimum values set at 50, 60 and 20 psi, respectively. Theeated nebulizer temperature was set at 500 ◦C. The ionspray volt-ge was set at 4000 V. The instrument response was optimised forfloxacin and the internal standard by infusing a solution of theompounds dissolved in mobile phase at a constant flow. The pauseime was set at 5 ms, the dwell time at 150 ms, and the collision gasN2) was set to medium (arbitrary values).

The AB Sciex API 3200 Q Trap mass spectrometer was oper-ted at unit resolution in the multiple reaction monitoring (MRM)ode, monitoring the transition of the protonated molecular ion/z 362.1 to the product ions at m/z 318.3 for ofloxacin, and the

d Biomedical Analysis 58 (2012) 177–181

protonated molecular ion m/z 376.1 to the product ions m/z 332.4for gatifloxacin. The instrument was interfaced with a computerrunning Applied Biosystems Analyst version 1.4.2 software.

2.7. Liquid chromatography

Chromatography was performed on a Phenomenex Luna, PFP(50 mm × 2.0 mm, 5 �m) analytical column. The mobile phase con-sisted of acetonitrile and water with 0.1% formic acid and wasdelivered with a gradient (10–70% acetonitrile over 5 min, broughtback to 10% in 0.1 min, and kept at 10% acetonitrile for another2.9 min) at a flow rate of 400 �l/min for 8 min. The analytical col-umn was kept in a column compartment at a constant temperatureof 20 ◦C. An Agilent 1200 series autosampler injected 5 �l onto theHPLC column. The injection needle was rinsed with mobile phase(acetonitrile and water with 0.1% formic acid (50:50, v/v)) for 10 susing the flush port wash station. The samples were cooled to 4 ◦Cwhile awaiting injection. A representative raw chromatogram atthe lower limit of quantification (LLOQ) for ofloxacin is presentedin Fig. 2.

2.8. Method validation

2.8.1. Calibration standards and quality controlsThe calibration curve for ofloxacin was validated by analysing

plasma QC samples in six fold at high, medium, and low concen-trations (16, 8 and 0.25 �g/ml, respectively) over a period of 3 daysto determine the intra- and inter-day accuracy and precision. TheQC values were interpolated from the calibration curve. The cal-ibration curve contained 10 different concentrations spanning aconcentration range of 0.078–20 �g/ml. Calibration curves wereconstructed using a weighted quadratic regression (1/concentra-tion) of the peak area ratio of ofloxacin to the internal standardversus nominal concentration.

2.8.2. RecoveryRecovery was evaluated at high, medium and low concen-

trations (16, 8 and 0.25 �g/ml, respectively). Blank plasma wasextracted and the dried samples reconstituted with mobile phaseand spiked with ofloxacin to generate the reference samples(containing matrix components). The high, medium and low QCstandards were used as the test samples. Recoveries were calcu-lated by comparing the peak areas of the extracted samples withthose of the reference samples.

2.8.3. Stock solution stabilityStock solutions of ofloxacin and the internal standard were pre-

pared in a mixture of acetonitrile and water (50:50, v/v). The teststocks were left at room temperature and the reference stocks at−70 ◦C for 6 h. Both the ofloxacin reference and test stocks werediluted with mobile phase (spiked with either the reference or teststock of internal standard) at a mid-level concentration (5 �g/ml)and were analysed with a valid calibration curve. In addition, refer-ence stocks stored at −70 ◦C for at least a year were diluted andcompared with freshly prepared stocks to determine long-termstability at −70 ◦C.

2.8.4. Freeze and thaw stabilityTo ascertain freeze–thaw stability, high (16 �g/ml) and low

(0.25 �g/ml) QC standards were frozen, put through three freeze(at −70 ◦C) and thaw (at room temperature) cycles and analysedagainst a valid calibration curve.

2.8.5. Benchtop stabilityTo establish benchtop stability, high (16 �g/ml) and low

(0.25 �g/ml) QC standards were frozen at −70 ◦C, left on the bench

S.A. Meredith et al. / Journal of Pharmaceutical and Biomedical Analysis 58 (2012) 177–181 179

XIC of

0.0

2000.0

4000.0

6000.0

8000.0

1.0e4

1.2e4

1.4e4

1.6e4

1.8e4

2.0e4

2.2e4

2.4e4

2.6e4

2.8e4

3.0e4

3.2e4

3.4e4

Inte

nsity

, cps

+MRM (4 pairs): 362.127/3

0.5 1.00

0

0

0

0

4

4

4

4

4

4

4

4

4

4

4

4

4

0.690.58 1.00.13

318.300 Da from Sample 1

0 1.5 2.01.3403 1.981.68

(B6) of 1044.wiff (Turbo Sp

2.5 3.03.04 32.782.572.49

pray)

3.5 4.03.19 3.66 3.95 44.07

4.5 5.0 54.64 5.13 5.215.02.55

5.5 6.0 6.56.6.295.81 6.155.60

Max. 6

7.0 7.560 7.257.05 7.316.66 7.5

60.0 cps.

54

plasm

ac

2

ipaa

2

Tinsm(cwtcbmi

2

ahmcp

Fig. 2. Chromatogram of a blank

t room temperature for 4 h, and were analysed against a validalibration curve.

.8.6. On-instrument stabilityA 72 h on-instrument stability evaluation of ofloxacin and the

nternal standard was performed. Six high QCs were extracted andooled, and analysed over four days. The samples were extractednd analysed on day 1, left on the autosampler and analysed againfter 24, 48 and 72 h.

.8.7. Matrix effects evaluationMatrix effects were evaluated both visually and quantitatively.

he publication by Matuszewski et al. was followed to evaluate thenfluence of co-eluting matrix components on ofloxacin and inter-al standard ionization [10]. Blank plasma from six different plasmaources was extracted. The dried samples were reconstituted withobile phase and spiked with ofloxacin at a high concentration

20 �g/ml), and diluted with mobile phase for a relatively low con-entration (2 �g/ml). The quantitative assessment of matrix effectsas obtained by comparing the peak area ratios. Visualization of

he matrix effects was achieved by post-column infusion [11]. Aontinuous infusion of ofloxacin and gatifloxacin was introducedy a Harvard infusion pump through a T-piece connector to theass spectrometer while plasma samples were injected. A change

n response would indicate ion suppression or enhancement.

.8.8. Haemolysis evaluationThe effect of haemolysis on analyte quantitation was evaluated

t 1% and 2% haemolysed blood present in plasma at a relatively

igh (16 �g/ml) and a relatively low (0.2 �g/ml) concentration. Theeasured concentration of the test samples were determined and

ompared with the measured concentration of the reference sam-les to calculate the overall accuracy.

Time, min

a sample with internal standard.

2.8.9. SpecificityThe high specificity of MS/MS detection prevents the detection

of compounds that do not produce the specific parent ion in theQ1 scan and the specific product ion in the Q3 scan. Specificity wasassessed by analysing extracts from six different plasma sources.

2.8.10. Effect of concomitant medication on the assayNine commonly administered anti-tuberculosis drugs, likely to

be co-administered with ofloxacin, were evaluated. The drugs;linezolid, terizidone, ethionamide, cycloserine, pyrazinamide,ethambutol, rifampicin, kanamycin and capreomycin, were spikedinto plasma at 250 ng/ml (test plasma). The reference plasma con-tained the same volume of solvent used to prepare the stocksolutions of the drugs. The effect on analyte quantitation was evalu-ated at high, medium and low concentrations (16, 8 and 0.25 �g/ml,respectively). The peak area of ofloxacin and the internal standardof the test samples were compared with that of the reference sam-ples to calculate the overall accuracy.

3. Results and discussion

The protein precipitation extraction method for ofloxacin per-formed well during validation. The calibration range was validatedbetween 0.078 �g/ml and 20 �g/ml. A quadratic regression, withpeak area ratio (drug/internal standard) against concentration with1/concentration (1/x) weighting, was fitted to the calibration curve.The combined accuracy and precision statistics of the quality con-trols (N = 18; high, medium and low) were between 97.1% and102.4%, and 2.8% and 5.2%, respectively. The percentage recovery for

ofloxacin was greater than 70% (N = 3) and reproducible over threelevels. According to FDA guidelines, recovery of the analyte neednot be 100%, but should be consistent, precise and reproducible[12]. At a relatively low, medium and high concentration the %

180 S.A. Meredith et al. / Journal of Pharmaceutical and Biomedical Analysis 58 (2012) 177–181

XIC of +MRM (4 pairs): 362.1/318.3 amu from Sample 1 (S1) of 1007.wiff (Turbo Spray) Max. 1273.3 cps.

0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5

Time, min

0

50

100

150

200

250

300

350

400

450

500

550

600

650

700

750

800

850

900

950

1000

1050

1100

1150

1200

1250

Inte

nsity

, cps

3.24

0.11

1.670.25 2.77-0.19 1.94 88.243.1!! - Noise -

S/N = 47.3

Peak Int.(Subt.)=1.3e+3

Ymax=2.7e+1 cps Ymin=0.0e+0 cps

calibration standard at LLOQ (0.078 �g/ml).

rr

aosNapHsdetwoo1opoto3

tcfaliieiea

Table 1Assessment of matrix effects of 6 different matrices at a high and low concentration.

H level (20 �g/ml) L Level (2 �g/ml)

Mean ratio of peak areas 121 169

Fig. 3. Chromatogram of an ofloxacin

ecovery was 76.5 (CV% = 3.5), 77.3 (CV% = 0.4) and 74.5 (CV% = 8.8),espectively. The CVs comply with these guidelines.

Long-term storage of the ofloxacin stock solution in a mixture ofcetonitrile and water (50:50, v/v) at −70 ◦C for a period longer thanne year was examined. The accuracy of the stored stock solutionamples compared to freshly prepared samples is 106.9 (CV% = 5.0,= 2). Ofloxacin stock solution showed acceptable stability after 6 h

t room temperature. The accuracy of the stock solution test sam-les compared to the reference samples is 90.1% (CV% = 5.9, N = 3).owever, stocks solutions were stored at −70 ◦C until required for

tandard and quality control preparation, and was spiked imme-iately into plasma after being thawed. Freeze–thaw stability wasvaluated over three freeze–thaw cycles on consecutive days athe high and low concentrations. The observed concentrationsere within 4% of the nominal concentrations, demonstrating that

floxacin is stable through three freeze–thaw cycles. The accuracyf ofloxacin benchtop stability over 4 h is 90.8% (CV% = 10, N = 5) at6 �g/ml and 96.6% (CV% = 2.7, N = 5) at 0.25 �g/ml, indicating thatfloxacin is stable at room temperature for up to 4 h. Lastly, sam-le extracts of ofloxacin and the internal standard are also stablen-instrument at 5 ◦C for up to 72 h. The calculated accuracies ofhe ratios (N = 5) were all well within 15% of the reference (ratiosn day 1). After 24 h (day 2) the accuracy was 99.5%, after 48 h (day) it was 100.4%, and after 72 h (day 4), 101.5%.

The effect of matrix components was evaluated both quan-itatively and visually at a relatively high and relatively lowoncentration. Quantitative assessment was performed using 6 dif-erent plasma sources. The coefficient of variation of the 6 peakreas of ofloxacin and the internal standard was 7.3% at the highevel and 5.1% at the low level (Table 1). In addition, no changen response for ofloxacin or the internal standard was observedn the post-column infusion experiment. No significant matrix

ffects were observed for ofloxacin or the internal standard, gat-floxacin. The effect of haemolysis on analyte quantitation wasvaluated at 1% and 2% haemolysed blood present in plasma. Theccuracy of the haemolysed sample (N = 6) at a relatively high

SD 8.81 8.69CV% 7.3 5.1

(16 �g/ml) experiment was 96.6% (CV% = 6.2) and 96.7 (CV% = 4.6)for 1% and 2% haemolysed blood, respectively. The accuracy at therelatively low (0.2 �g/ml) experiment was 100.9% (CV% = 6.2) and101.2 (CV% = 1.5) for 1% and 2% haemolysed blood, respectively. Thepresence of haemolysed blood was found to have no impact onofloxacin quantitation. Similarly, no significant effect on analytequantitation was observed in the presence of other drugs likely tobe co-administered with ofloxacin. The accuracy of the test samplecontaining the nine drugs (N = 3) at a relatively high concentra-tion (16 �g/ml) was 91.8% (CV% = 5.2), at a mid-level concentration(8 �g/ml) was 91.1% (CV% = 6.7), and at a relatively low concentra-tion was 93.2% (CV% = 7.6).

Due to the high specificity of MS/MS detection, no interferingor late eluting peaks were found when analysing blank plasma(Fig. 2). This is confirmed with the analysis of extracts from six dif-ferent plasma sources. The lower limit of quantification (LLOQ) forthe method, when injecting 5 �l onto the column, is 0.078 �g/ml(Fig. 3). The signal to noise at LLOQ was well above the minimuminternational accepted criteria (S/N > 5). The accuracy for the LLOQ(N = 6) was 95.7% (CV% = 6.6), well within acceptable limits (accu-racy within 80–120%, precision below 20%). If necessary, sensitivityof the method could be improved by raising the injection volume.

4. Application to clinical pharmacokinetics studies

The validated method performed well during sample analysis ofclinical samples generated during a clinical study conducted at DPMarais Hospital, Cape Town, South Africa. The objective of the study

S.A. Meredith et al. / Journal of Pharmaceutical an

woMiroa

o(L

5

Lt0apwtgaw

[

[11] R. Bonfiglio, R.C. King, T.V. Olah, K. Merkle, The effects of sample preparation

Fig. 4. Concentration vs. time profile of ofloxacin (N = 16).

as to characterize the population pharmacokinetics of a numberf second line anti-tuberculosis drugs utilized in the treatment ofDR-TB. The assay was applied to determine the levels of ofloxacin

n human plasma following oral administration of 800 mg. A rep-esentative mean concentration versus time profile (up to 12 h) offloxacin for this pharmacokinetic study of anti-tuberculosis ther-py in patients with MDR-TB is presented in Fig. 4 (16 patients).

The precision (total-assay coefficients of variation; CV %) forfloxacin during sample analysis were less than 8% at high16 �g/ml), medium (8 �g/ml) and low (0.25 �g/ml) QC levels. TheLOQ was 0.078 �g/ml.

. Conclusion

A simple protein-precipitation extraction method coupled withC–MS/MS detection has been developed and fully validated forhe determination of ofloxacin, with a quantification limit of.078 �g/ml. This method is specific, sensitive and reproduciblend has been successfully used to quantify ofloxacin in a clinicalharmacokinetic study. A number of methods using HPLC coupledith UV or fluorescence detection [3,5–7] allow for the quan-

ification of ofloxacin, but require plasma volumes equal to orreater than 200 �l. The LC–MS/MS method described here usesmuch lower plasma volume (20 �l) and injection volume (5 �l),ith good sensitivity at the LLOQ. The lower plasma volume is

[

d Biomedical Analysis 58 (2012) 177–181 181

important for pharmacokinetic studies and therapeutic drug mon-itoring in children, and the low injection volume offers thepossibility of enhancing sensitivity. This method can be usedfor pharmacokinetic studies and therapeutic drug monitoring inpatients, including children, with MDR-TB.

Acknowledgements

I would like to acknowledge Helen McIlleron (Principal Investi-gator), Emmanuel Chigutsa (PK Investigator) and Rae Taylor (StudyCo-ordinator) of the Population Pharmacokinetics of second lineanti-tuberculosis drugs for multi drug resistant tuberculosis clini-cal trial in Cape Town, South Africa, for allowing the authors to usetheir analytical results to further validate the developed assay.

References

[1] H. Tomioka, Prospects for the development of new antimycrobacterial drugs,J. Infect. Chemother. 6 (2000) 8–20.

[2] S.M. Hwang, D.D. Kim, S.J. Chung, C.K. Shim, Delivery of ofloxacin to the lungand alveolar macrophages via hyaluronan microspheres for the treatment oftuberculosis, J. Control. Release 129 (2008) 100–106.

[3] J. Macek, E. Ptácek, Determination of Ofloxacin in human plasma using high-performance liquid chromatography and fluorescence detection, J. Chromatogr.B: Biomed. Sci. Appl. 673 (1995) 316–319.

[4] D.C. Hooper, J.S. Wolfson, Fluoroquinolone antimicrobial agents, Clin. Micro-biol. Rev. 2 (1989) 378–424.

[5] C. Wongsinsup, W. Taesotikul, S. Kaewvichit, S. Sangsrijan, S. Sangsrijan,Simple extraction and determination of ofloxacin in human plasma by high-performance liquid chromatography with fluorescence detector, CMU Nat. Sci.8 (2009) 165–174.

[6] M. Amini, Kh. Abdi, M. Darabi, A. Shafiee, Determination of ofloxacin in plasmaby HPLC with UV detection, J. Appl. Sci. 5 (2005) 1655–1657.

[7] D.K. Xu, A.Z. Ding, Y.S. Yuan, Y. Diao, Determination of ofloxacin in humanplasma and studies of its pharmacokinetics using HPLC method, Yao Xue XueBao 27 (1992) 462–466.

[8] R. Kauffman, G. Kearns, Pharmacokinetic studies in pediatric patients, clinicaland ethical considerations, Clin. Pharmacokinet. 23 (1992) 10–29.

[9] J.L. Veuthey, S. Souverain, S. Rudaz, Column-switching procedures for thefast analysis of drugs in biologic samples, Ther. Drug Monit. 26 (2004)161–166.

10] B.K. Matuszewski, M.L. Constanzer, C.M. Chavez-Eng, Strategies for theassessment of matrix effect in quantitative bioanalytical methods based onHPLC–MS/MS, Anal. Chem. 75 (2003) 3019–3030.

methods on the variability of the electrospray ionization response for modeldrug compounds, Rapid Commun. Mass Spectrom. 13 (1999) 1175–1185.

12] Guidance for Industry Bioanalytical Method Validation, Department of Healthand Human Services, Food and Drug Administration, Rockville, USA, 2001.