HPLC-UV analysis of thymidine and deoxyuridine in plasma of patients with thymidine phosphorylase deficiency

Journal of Chromatography B, 949–950 (2014) 58–62

Contents lists available at ScienceDirect

Journal of Chromatography B journal homepage: www.elsevier.com/locate/chromb

Short Communication

HPLC-UV analysis of thymidine and deoxyuridine in plasma of patients with thymidine phosphorylase deficiency Susan Mohamed a , Leonardo Caporali a , Roberto De Giorgio c , Valerio Carelli a,b , Manuela Contin a,b,∗ a

IRCCS-ISNB Institute of Neurological Sciences of Bologna, Bologna, Italy Department of Biomedical and Neuromotor Sciences (DIBINEM), University of Bologna, Bologna, Italy c Department of Medical and Surgical Sciences/Digestive Diseases and Internal Medicine, St. Orsola-Malpighi Hospital, University of Bologna, Bologna, Italy b

a r t i c l e

i n f o

Article history: Received 27 August 2013 Received in revised form 8 December 2013 Accepted 2 January 2014 Available online 8 January 2014 Keywords: Thymidine Deoxyuridine Mitochondrial neurogastrointestinal encephalomyopathy High performance liquid chromatography UV detection

a b s t r a c t We present a simple, fast and validated method for the determination of the two nucleosides thymidine (dThd) and deoxyuridine (dUrd) in plasma of patients with symptoms suggestive of mitochondrial neurogastrointestinal encephalomyopathy (MNGIE), using high performance liquid chromatography coupled with ultraviolet spectrophotometric detection (HPLC-UV). Plasma sample (100 ␮L) pretreatment was based on simple deproteinization by 1.2 M perchloric acid, using theophylline as internal standard (I.S.). HPLC-UV analysis was carried out on a Synergi 4 ␮m Hydro-RP, 150 × 4 mm I.D. column, at room temperature. The mobile phase was a mixture of potassium dihydrogen phosphate buffer (20 mM, pH 4.5) and acetonitrile (95:5, v/v), at an isocratic flow rate of 0.7 mL/min. The UV detector was set at 267 nm. The chromatographic run lasted 19 min. Similar pyrimidine nucleotides and nucleosides do not interfere with the assay. Calibration curves were linear for both dThd and dUrd over a range of 0.5 to 5.0 ␮g/mL. The limit of quantitation was 0.5 ␮g/mL for both nucleosides and the absolute recovery was >90% for dThd, dUrd and the I.S. Both intra- and inter-assay precision and accuracy were lower than 10% at all tested concentrations. The proposed method was successfully applied to measure plasma concentrations of dThd and dUrd in two MNGIE patients. This assay simplifies both plasma pretreatment and chromatographic conditions of previously reported procedures and describes the first validated method for the determination of the two nucleotides in human plasma. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Mitochondrial neurogastrointestinal encephalomyopathy (MNGIE) is a rare autosomal recessive disorder clinically characterized by ptosis and ophthalmoparesis, gastrointestinal dysmotility, peripheral neuropathy, leukoencephalopathy, and mitochondrial dysfunctions. MNGIE is due to mutations in TYMP gene, which encodes thymidine phosphorylase (TP) [1,2]. The prevalence of MNGIE is currently unknown: in the last 12 years, since the identification of TYMP mutations as causative of MNGIE, the largest cohort of 102 patients was recently published [3] and a few other scattered reports are available from literature suggesting that the total number of patients worldwide is possibly less than 200.

∗ Corresponding author at: IRCCS, Institute of Neurological Sciences of Bologna, Department of Biomedical and Neuromotor Sciences, University of Bologna, Via Altura 3, 40139 Bologna, Italy. Tel.:+39 051 4966752; fax: +39 051 4966208. E-mail address: [email protected] (M. Contin). 1570-0232/$ – see front matter © 2014 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jchromb.2014.01.003

TYMP mutations cause loss of function of the enzyme, leading to an unbalanced metabolism of thymidine (dThd) and deoxyuridine (dUrd) and to their systemic accumulation in body fluids and tissues. Reported plasma concentrations of dThd and dUrd in MNGIE patients range from 0.9 to 4.3 ␮g/mL and 1.25 to 5.6 ␮g/mL, respectively [4–6], while they are undetectable (<0.01 ␮g/mL) in healthy controls. Analyzing plasma dThd and dUrd concentration is the easiest way to test for TP dysfunction and support TP activity assay [4–6]. In literature only a few papers described methodologically the determination of dThd and dUrd in plasma of patients with altered thymidine metabolism, based on HPLC-UV detection [4,7], or tandem mass spectrometry [8], involving time-consuming chromatographic gradient elution procedures to eliminate interferences [4,7], complex sample pre-treatment [7], and lacking analytical validation. Here we propose a validated HPLC-UV assay for the determination of dThd and dUrd in plasma of MNGIE patients, which simplifies both plasma pretreatment and chromatographic conditions of previous methods.

S. Mohamed et al. / J. Chromatogr. B 949–950 (2014) 58–62

2. Experimental 2.1. Reagents and standards

59

Table 1 List of compounds checked for thymidine and deoxyuridine assay interference. Compound

Concentration (␮g/mL)

Retention time (min)

Adenosine Cytidine Cytosine Guanosine Inosine Thymine Tryptophan Uric acid Uridine Xanthine

5 5 5 5 5 5 5 5 5 5

8.0 3.2 n.d 4.3 4.2 4.5 11.1 n.d n.d n.d

Thymidine, deoxyuridine, the internal standard (I.S.) theophylline and thymine, cytosine, uric acid, cytidine, uridine, xanthine, inosine, guanosine, tryptophan, adenosine were purchased from Sigma Aldrich (St. Louis, MO, USA). HPLC grade acetonitrile, 12 M perchloric acid and potassium dihydrogen phosphate were purchased from Merck (Darmstadt, Germany). Ultrapure water was obtained from a MilliQ Gradient A10 apparatus (Merck Millipore, Darmstadt, Germany). Frozen, drug-free plasma (blank plasma) for calibrators preparation was obtained from the blood bank of the Maggiore Hospital of Bologna, stored at −20 ◦ C and thawed at room temperature before use. Stock solutions (1 mg/mL) and subsequent dilutions (500, 250, 125 and 50 ␮g/mL, working solutions) of dThd and dUrd were prepared by dissolving pure nucleosides in 0.1 M perchloric acid. Internal standard stock solution (1 mg/mL) was prepared by dissolving 10 mg of theophylline in 10 mL ultrapure water. All solutions were prepared monthly and stored at 4 ◦ C. Calibrators (calib, 1–4) of 0.5, 1.25, 2.5, 5.0 ␮g/mL for both dThd and dUrd were prepared by pipetting 10 ␮L of the nucleosides working solutions (50 ␮g/mL, calib 1, 125 ␮g/mL, calib 2, 250 ␮g/mL, calib 3, and 500 ␮g/mL, calib 4) to 1 mL aliquots of blank plasma. Plasma calibrators were prepared fresh for each batch and then treated exactly as subject specimens.

for 10 min at 4 ◦ C. Plasma was separated in 1-mL aliquots and stored at −80 ◦ C until analysis (within 2 months). The study was approved by the Ethics Committee of the Bologna University Hospital Authority St. Orsola-Malpighi Polyclinic (protocol number 731/2013). Written informed consent was obtained from each subject. One hundred microliters plasma aliquots (calibrators, healthy volunteers or patient samples) were spiked with 10 ␮L of I.S. solution (100 ␮g/mL), deproteinized by addition of 100 ␮L 1.2 M perchloric acid, vortexed for 30 s, and diluted with 800 ␮L of ultrapure water. The samples were then centrifuged at 2500 × g at 4 ◦ C for 10 min. Twenty microliters of the clean upper layer were injected into the chromatographic system. HPLC injection and elution were performed at thermostated room temperature (23 ◦ C).

2.2. Chromatographic apparatus and conditions

2.5. Method validation

The HPLC system consisted of a Series 200 liquid chromatograph, a Series 200 UV/VIS spectrophotometric detector, set at 267 nm, and a Series 200 autosampler connected by a model 600 link chromatography interface to the TotalChrom chromatography workstation. All the equipment was purchased from Perkin Elmer, Norwalk, CA, USA. The chromatographic separation was performed with a Synergi 4 ␮m Hydro-RP, 150 × 4.6 mm I.D. column (Phenomenex, Torrance, CA, USA), protected by a C18 Securityguard precolumn (Phenomenex) and a graphite filter (ESA, Chelmsford, MA, USA), fitted between the autosampler and precolumn. The mobile phase was a mixture of potassium dihydrogen phosphate buffer (20 mM, pH 4.5), filtered through a 0.22 ␮m membrane filter (GS type, Millipore, Darmstadt, Germany) and acetonitrile (95:5, v/v). The flow rate was set at 0.7 mL/min. Column cleaning procedures between batches included a first rinse with 10 column volumes of water and acetonitrile (95:5, v/v), followed by a second flush with 10 column volumes of a mixture of acetonitrile and water (50:50, v/v) and concluded with 10 column volumes of water and acetonitrile (35:65, v/v).

Calibration curves for dThd and dUrd were run on each analysis day (n = 6) for two months. The analyte to I.S. peak area ratios were plotted against dThd and dUrd matched concentration added to the blank plasma. The calibration curves were calculated by the least square method. Linearity was assessed by determining the coefficient of correlation (r) of the points of the curves. Plasma concentrations of dThd and dUrd were expressed in ␮g/mL (conversion factors to ␮mol/L, 4.13 for dThd and 4.38 for dUrd). For assay precision and accuracy assessment, blank plasma pools were spiked using suitable volumes of the same dThd and dUrd working solutions used for the calibrators, to yield three concentrations (i.e., 0.5, 2.5, 5.0 ␮g/mL) corresponding to the lower, middle and upper points of the calibration curve (Quality Controls, QCs). The precision of the method was assessed by determining the relative standard deviation (R.S.D. = 100 × SD/mean) at the three plasma dThd and dUrd concentrations within the same analysis (n = 6, intraday precision) and in triplicate over a series of six analysis (n = 18, interday precision). The accuracy of the method was determined by comparing the means of dThd and dUrd calculated concentrations in the abovementioned QCs with the nominal concentrations (percentage differences), within the same analysis (n = 6, intraday accuracy) and in triplicate over a series of six analysis (n = 18, interday accuracy). The absolute recovery of dThd, dUrd and I.S. was calculated as the ratio of the two nucleosides peak area from deproteinized blank plasma spiked with dThd and dUrd, at the three concentrations above, or with the I.S. (10 ␮g/mL) to the peak area obtained from the injection of dThd, dUrd and I.S. standard solutions, at the same theoretical concentrations, reconstituted in 0.1 M perchloric acid, over a series of six analyses. The lower limit of quantitation (LOQ) was defined as the lowest concentration of the calibration curves matched to a dThd and dUrd signal-to-noise ratio 10:1, with an associated R.S.D. and inaccuracy <20% [10]. The precision and accuracy at the LOQ were determined

2.3. Method specificity Standard solutions of potentially co-eluting compounds [9], including pyrimidine nucleotides (thymine, cytosine, uric acid) and nucleosides (cytidine, uridine) were injected to check for possible interferences (Table 1). In addition, a series of 11 plasma control samples from healthy volunteers (age 25–43 years, 7 females) were tested for endogenous interferences. 2.4. Blood sampling and plasma processing Venous blood samples (5 mL) drawn from healthy volunteers and two MNGIE patients were transferred into heparinized tubes (8 IU heparin/mL blood) and immediately centrifuged at 1500 × g

n.d., not detectable.

60

S. Mohamed et al. / J. Chromatogr. B 949–950 (2014) 58–62

Fig. 1. Chromatograms obtained by injecting 20 ␮L of (a) deproteinized control plasma; (b) deproteinized blank plasma spiked with: deoxyuridine (2.5 ␮g/mL), (1) inosine (5.0 ␮g/mL), (2) guanosine (5.0 ␮g/mL), (3) thymine (5.0 ␮g/mL), thymidine (2.5 ␮g/mL), (4) adenosine (5.0 ␮g/mL), (5) tryptophan, (5.0 ␮g/mL), and I.S., (10.0 ␮g/mL); (c) deproteinized blank plasma spiked with dUrd, 2.5 ␮g/mL, dThd, 2.5 ␮g/mL and I.S., 10.0 ␮g/mL; (d) deproteinized plasma specimen of MNGIE patient 1: dUrd, 4.5 ␮g/mL and dThd, 3.6 ␮g/mL. dUrd, deoxyuridine; dThd, thymidine; I.S., internal standard.

both intraday (n = 6) and interday (triplicate samples over six analysis, n = 18) [10]. The lower limit of detection (LOD) was determined in triplicate by comparing measured signals from blank plasma samples spiked

with known low concentrations of dThd and dUrd with those of blank plasma and calculated as 3 times the baseline noise [10]. During the development and optimization phase of the method its robustness was checked with respect to variations in mobile

S. Mohamed et al. / J. Chromatogr. B 949–950 (2014) 58–62

61

Table 2 Precision and accuracy of thymidine and deoxyuridine assay. Amount added to blank plasma (␮g/mL)

dThd 0.5 (LOQ) 2.5 5.0 dUrd 0.5 (LOQ) 2.5 5.0

Intraday (n = 6)

Interday (n = 18)

Calculated concentration (mean ± SD) (␮g/mL)

Precision (RSD%)

Accuracy (%)

Calculated concentration (mean ± SD) (␮g/mL)

Precision (RSD%)

Accuracy (%)

0.49 ± 0.03 2.41 ± 0.08 4.90 ± 0.21 0.48 ± 0.04 2.64 ± 0.11 4.94 ± 0.16

6.1 3.3 4.3 8.3 4.2 3.2

−2.0 −3.6 −1.8 −4.0 5.6 −1.0

0.54 ± 0.05 2.48 ± 0.15 4.97 ± 0.19 0.52 ± 0.05 2.53 ± 0.09 5.02 ± 0.21

9.2 6.0 3.9 9.6 3.5 4.2

8.0 −0.8 −0.5 4.0 1.2 0.4

Precision (RSD%) = 100 × SD/mean; Accuracy (%) = 100 × (mean concentration found − known concentration)/known concentration); Interday (n = 18) = triplicate samples, over a series of six analyses on different days; LOQ = limit of quantitation.

phase flow rate (from 0.7 to 1.0 mL/min) [11]. Precision and accuracy of low, medium and high QCs were determined in triplicate at each flow rate. In order to assess the autosampler stability of processed samples, low and high QCs were freshly analyzed in triplicate and stored in autosampler at room temperature (23 ◦ C) for 24 h. QC samples were analyzed against a calibration curve, obtained from freshly spiked calibration standards, and the obtained concentrations were compared to the nominal concentrations. Processed QCs were considered stable if their mean concentration was within ±15% of the nominal concentration [12].

(±0.00413) + 0.08742(±0.01570)x, r = 0.9990(±0.0006) for and y = 0.00173(±0.00619) + 0.09070(±0.00655)x, dThd r = 0.9990(±0.00037) for dUrd, where x is dThd or dUrd concentration, expressed in ␮g/mL, y is the analyte to I.S. peak area ratio, expressed in arbitrary area units and r is the correlation coefficient. The results of precision and accuracy analysis for dThd and dUrd are reported in Table 2. Both intra and interassay precision and accuracy were <±10% for the whole concentration range. The LOQ was set at 0.5 ␮g/mL for both dThd and dUrd. The LOD was 0.12 ␮g/mL for both analytes. Mean ± S.D. absolute recovery for dThd was 91.5 ± 8.9% at a plasma nucleoside concentration of 0.5 ␮g/mL, 90.6 ± 0.8% at 2.5 ␮g/mL and 93.8 ± 0.9% at 5.0 ␮g/mL. Recovery for dUrd was 89.2 ± 4.4% at a plasma nucleoside concentration of 0.5 ␮g/mL, 93.1 ± 2.1% at 2.5 ␮g/mL and 95.2 ± 7.2% at 5.0 ␮g/mL; recovery for the I.S. (10 ␮g/mL) was 96.9 ± 6.6%. The results of robustness analyses with respect to variations in mobile phase flow rate are reported in Table 3. Precision and accuracy of dThd and dUrd assay were within ±10% at both flow rates. As far as 24-hour processed samples autosampler stability is concerned, calculated mean ± S.D. plasma concentrations of QC samples were 0.45 ± 0.04 ␮g/mL for dUrd and 0.51 ± 0.05 ␮g/mL for dThd (low QC); 4.96 ± 0.22 ␮g/mL and 5.15 ± 0.32 ␮g/mL for dUrd and dThd, respectively (high QC). Mean plasma concentrations of dUrd and dThd resulted within ±10% of the nominal concentrations for both QCs. The method we developed was applied to the analysis of dUrd and dThd in plasma samples of two molecularly confirmed MNGIE patients, a 30-year-old woman (patient 1) and a 38-yearold woman (patient 2). Measured plasma concentrations were 4.5 ␮g/mL for dUrd and 3.6 ␮g/mL for dThd in patient 1 (Fig. 1d);

3. Results and discussion The mixture of mobile phase (potassium dihydrogen phosphate buffer, 20 mM, pH 4.5 and acetonitrile, 95:5, v/v) combined with a Synergi 4 ␮m Hydro-RP column 150 × 4.6 mm at a flow rate of 0.7 mL/min allowed a good separation of dThd and dUrd with mean ± S.D. (n = 6) of 4.0 ± 0.1 min for dUrd, 6.9 ± 0.1 min for dThd and 18.9 ± 0.6 min for the I.S. No endogenous interferences were found either in blank plasma pools or in any of the 11 healthy volunteers’ plasma specimens (Fig. 1a). Elution times of similar compounds tested over a 30-minute run are reported in Table 1. A chromatogram obtained by injecting a deproteinized blank plasma spiked with a mixture of inosine, guanosine, thymine, adenosine, tryptophan, plus dUrd, dThd and the I.S. is depicted in Fig. 1b. As it can be seen, dUrd is sufficiently resolved from contiguous inosine and guanosine and no other compound turns out to potentially interfere in the analysis. Calibration curves showed a linear and reproducible correlation between dThd and dUrd plasma concentrations and matched analyte to I.S. peak area ratios. Equations (mean ± S.D., n = 6) of the regression lines were: y = 0.00150

Table 3 Robustness of thymidine and deoxyuridine assay with respect to variations in mobile phase flow rate. Mobile phase flow rate Amount added to blank plasma (␮g/mL)

dThd 0.5 2.5 5.0 dUrd 0.5 2.5 5.0

1.0 mL/min (n = 9)

0.7 mL/min (n = 9)

Calculated concentration (mean ±SD) (␮g/mL)

Precision (RSD%)

Accuracy (%)

Calculated concentration (mea n ±SD) (␮g/mL)

Precision (RSD%)

Accuracy (%)

0.54 ± 0.05 2.71 ± 0.16 5.28 ± 0.24 0.53 ± 0.04 2.58 ± 0.16 5.36 ± 0.26

9.2 5.9 4.5 7.5 6.2 4.8

8.0 8.4 5.6 6.0 3.2 7.2

0.53 ± 0.03 2.43 ± 0.04 5.01 ± 0.14 0.50 ± 0.05 2.45 ± 0.06 4.94 ± 0.18

5.7 1.6 2.8 10 2.4 3.6

6.0 −2.8 0.2 0.0 −2.0 −1.2

Precision (RSD%) = 100 × SD/mean; Accuracy (%) = 100 × (mean concentration found − known concentration)/known concentration); n = 9, triplicate analyses for each low, medium and high QC sample.

62

S. Mohamed et al. / J. Chromatogr. B 949–950 (2014) 58–62

Table 4 Main characteristics of published analytical methods for the quantitative determination of thymidine and deoxyuridine in human biological fluids. Technique

Sample

Sample preparation

Elution

Run time

Method validation

Ref.

HPLC/UV (fixed ␭) HPLC/UV (variable ␭)

Plasma, urine Plasma

Gradient Gradient

110min 18min

No No

[4] [7]

HPLC/ESI-MS/MS

Plasma, urine

PP (HClO4 ) Two steps PP: 1. HClO4 2. K2 CO3 Diluition (H2 O/CH3 COONH4 )

Yes (dThd)

[8]

Isocratic

8.5 min

PP, protein precipitation; HClO4 , perchloric acid; K2 CO3 , potassium carbonate; CH3 COONH4 , ammonium acetate; ESI-MS/MS, electrospray ionization tandem mass spectrometry; dThd, thymidine.

1.25 ␮g/mL for dUrd and 0.6 ␮g/mL for dThd in patient 2. These results are in line with reported plasma concentration ranges for the two nucleosides in MNGIE patients [4–6]. The two nucleosides were undetectable in any of the healthy control plasma specimens. The long-term stability of stored plasma samples for dThd and dUrd quantitation is missing in the previous literature [4,7,8]. We re-analyzed the plasma of patient 1 after a 5-month storage at −80 ◦ C, obtaining values of 4.5 ␮g/mL for dUrd and 3.7 ␮g/mL for dThd, which are superimposable to those measured after 1-month storage. This procedure is the first validated method for the determination of both dThd and dUrd in plasma of MNGIE patients. Compared with the analytical methods reported so far (Table 4), this assay significantly simplifies both chromatographic conditions, omitting time-consuming, up to 110 minute-long gradient elution patterns [4,7], and sample purification, avoiding multiple-step sample pretreatment [7], thus, reducing the risks of analytical errors. The method quantitation range for the two analytes proved to be adequate for diagnostic purposes in MNGIE patients [4–6]. The LOQ of the assay was set at the lowest plasma concentration values of reported ranges for dThd and dUrd in MNGIE [4–6]. The statistical validation shows good intra- and inter-assay precision and accuracy within the whole concentration range for the two nucleosides, an optimal extraction efficiency and a good robustness of the method at the tested variations in mobile phase flow rate. Autosampler stored processed samples proved to be stable for at least 24 h. 4. Conclusion The proposed method proved to possess adequate specificity, sensitivity, accuracy and precision for a reliable determination of dThd and dUrd in plasma of patients with symptoms suggestive of MNGIE, confirming the drastic reduction of the enzymatic activity found in buffy coats of patients [4–6] and is validated according to recommended guidelines [10]. The minimal sample pretreatment allows patient samples to be processed in a short time. Moreover, the simple reversed-phase HPLC-UV chromatographic apparatus

means the method can be adopted even in laboratories lacking sophisticated analytical equipment. Acknowledgments The skillful technical assistance of Carmina Candela, Anita Calisti and Monica Balboni is gratefully acknowledged. Cecilia Baroncini edited the English text. We also thank Rita Rinaldi and Loris Pironi from St. OrsolaMalpighi Hospital in Bologna (Italy) and Costanza Lamperti and Massimo Zeviani from National Neurological Institute Carlo Besta in Milan (Italy) for referring the samples. We are also grateful to the MNGIE patients and their families for participating to this study. This project received support from the Italian Ministry of Health: Ricerca Finalizzata 2009 RF-2009-1492481 (to RDG). References [1] I. Nishino, A. Spinazzola, M. Hirano, Science 283 (1999) 689–692. [2] A. Spinazzola, R. Martí, I. Nishino, A.L. Andreu, A. Naini, S. Tadesse, I. Pela, E. Zammarchi, M.A. Donati, J.A. Oliver, M. Hirano, J. Biol. Chem. 277 (2002) 4128–4133. [3] C. Garone, S. Tadesse, M. Hirano, Brain 134 (2011) 3326–3332. [4] R. Martí, Y. Nishigaki, M. Hirano, Biochem. Biophys. Res. Commun. 303 (2003) 14–18. [5] R. Martí, A. Spinazzola, S. Tadesse, I. Nishino, Y. Nishigaki, M. Hirano, Clin. Chem. 50 (2004) 120–124. [6] M.L. Valentino, R. Martí, S. Tadesse, L.C. López, J.L. Manes, J. Lyzak, A. Hahn, V. Carelli, M. Hirano, FEBS Lett. 581 (2007) 3410–3414. [7] R. Massa, A. Tessa, M. Margollicci, V. Micheli, A. Romigi, G. Tozzi, C. Terracciano, F. Piemonte, G. Bernardi, F.M. Santorelli, Neuromuscul. Disord. 19 (2009) 837–840. [8] G. la Marca, S. Malvagia, B. Casetta, E. Pasquini, I. Pela, M. Hirano, M.A. Donati, E. Zammarchi, J. Mass. Spectrom. 41 (2006) 586–592. [9] R. Martí, L.C. Lopez, M. Hirano, in: L-J.C. Wong (Ed.), Mitochondrial Disorders, Biochemical and Molecular Analysis, Humana Press, Springer, New York, 2012, pp. 121–133. [10] ICH Harmonised Tripartite Guideline prepared within the Third International Conference on Harmonisation of Technical Requirements for the Registration of Pharmaceuticals for Human Use (ICH), Validation of Analytical Procedures: Methodology, 1996, pp. 1–13. [11] Y. Vander Heyden, A. Nijhuis, J. Smeyers-Verbeke, B.G. Vandeginste, D.L. Massar, J. Pharm. Biomed. Anal. 24 (2001) 723–753. [12] EMEA/CHMP/EWP/192217/2009, Committee for Medicinal Products for Human Use (CHMP), Guideline on Bioanalytical Method Validation, 2011, pp. 1–22.