2713

Electrophoresis 2015, 36, 2713–2719

Lee Yien Thang1,2 Shafinaz Shahir3 Hong Heng See1,2,4 1 Centre

for Sustainable Nanomaterials, Ibnu Sina Institute for Scientific and Industrial Research, Universiti Teknologi Malaysia, Johor, Malaysia 2 Department of Chemistry, Faculty of Science, Universiti Teknologi Malaysia, Johor, Malaysia 3 Department of Biosciences and Health Sciences, Faculty of Biosciences and Medical Engineering, Universiti Teknologi Malaysia, Johor, Malaysia 4 Australian Centre for Research on Separation Science, School of Physical Sciences – Chemistry, University of Tasmania, Tasmania, Australia

Received March 28, 2015 Revised June 16, 2015 Accepted July 6, 2015

Research Article

Determination of tamoxifen and its metabolites in human plasma by nonaqueous capillary electrophoresis with contactless conductivity detection A new approach for the quantification of tamoxifen and its metabolites 4hydroxytamoxifen, N-desmethyltamoxifen, and 4-hydroxy-N-desmethyltamoxifen (endoxifen) in human plasma samples using NACE coupled with contactless conductivity detection (C4 D) is presented. The buffer system employed consisted of 7.5 mM deoxycholic acid sodium salt, 15 mM acetic acid, and 1 mM 18-crown-6 in 100% methanol. The complete separation of all targeted compounds (including endoxifen racemate) could be achieved within 6 min under optimized conditions. The proposed method was validated and showed good linearity in the range from 100 to 5000 ng/mL with correlation coefficients between 0.9922 and 0.9973, LODs in the range of 25–40 ng/mL, and acceptable reproducibility of the peak area (intraday RSD 2.2–3.1%, n = 4; interday (3 days) RSD 6.0–8.8%, n = 4). The developed method was successfully demonstrated for the quantification of tamoxifen and its metabolites in human plasma samples collected from breast cancer patients undertaking tamoxifen treatment. Keywords: Breast cancer / Contactless conductivity detection / Nonaqueous capillary electrophoresis / Plasma / Tamoxifen DOI 10.1002/elps.201500164

1 Introduction Tamoxifen, which is a selective estrogen receptor modulator, is the most widely used anti-estrogen drug for the treatment of all stages of estrogen receptor positive breast cancer. It also acts as a chemopreventive agent for women with a high risk of developing breast cancer as well as an adjuvant for the prevention of recurrence after the removal of primary tumors [1–3]. Studies report that 30–50% of patients under longterm treatment with tamoxifen experienced relapse during the adjuvant treatment. Relapse could be caused by tumor- or patient-specific factors, such as differences in gene expression within the tumor cells, germline genetic variability, patient behavior, and other factors [4]. Hence, a monitoring assay for tamoxifen and its major metabolites is crucial in guiding treatment decisions. Tamoxifen is a pro-drug and can be metabolically activated to 4-hydroxytamoxifen and 4-hydroxy-N-

Correspondence: Dr. Hong Heng See, Centre for Sustainable Nanomaterials, Ibnu Sina Institute for Scientific and Industrial Research, Universiti Teknologi Malaysia, 81310 UTM Johor Bahru, Johor, Malaysia. E-mail: [email protected] Fax: +60-7-553-6080

Abbreviations: CYP2D6, cytochrome P450 2D6; EtOH, absolute ethanol; MeOH, methanol; NaDCHA, deoxycholic acid sodium salt; NMF, N-methylformamide  C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

desmethyltamoxifen (endoxifen) by the cytochrome P450 2D6 (CYP2D6). Alternatively, the tamoxifen can be metabolically routed via N-desmethyltamoxifen to endoxifen [5–7]. The 4-hydroxylated metabolites exhibit an increased binding affinity for estrogen receptor alpha, and thus suppress cell growth more effectively than tamoxifen [4]. The plasma levels of endoxifen in patients with functional CYP2D6 frequently exceed the levels of 4-hydroxytamoxifen [8]. Recent studies have shown that endoxifen is responsible for the majority of tamoxifen’s anti-estrogenic effects as the average concentrations of endoxifen are six times higher than 4-hydroxytamoxifen [7,9]. Wu et al. [10] emphasized that endoxifen is a potent antiestrogen that functions by targeting estrogen receptor alpha for degradation via the proteasome in breast cancer cells. In fact, endoxifen also blocks estrogen receptor alpha transcriptional activity and inhibits estrogen-induced breast cancer cell proliferation. Women with genetically impaired CYP2D6 have a higher risk of breast cancer recurrence as it significantly reduces the endoxifen concentration in the body. Several analytical techniques have been reported for the determination of tamoxifen and its metabolites. However, only a few papers have been published on the simultaneous determination of tamoxifen and endoxifen [11–15]. Mihailescu et al. [16] identified tamoxifen and its metabolites using GC-MS. GC has disadvantages in sample handling, as it requires sample derivatization. HPLC with UV absorbance [12, 13, 17, 18] is commonly employed in detection of tamoxifen and its metabolites as these analytes contain

www.electrophoresis-journal.com

2714

L. Y. Thang et al.

chromophores. In order to achieve higher detection sensitivity and separation efficiency, Arellano et al. [11] demonstrated the separation of tamoxifen and its metabolites using UPLC with tandem MS detection. Peris-Vicente et al. [15] developed a micellar LC-based method with fluorescence detection to quantify tamoxifen and its metabolites. The micellar media were reported to have advantages of reduced cost, simple preparation, and lower solvent consumption. NACE with UV absorbance also has been reported for the separation of tamoxifen and its metabolites [19–22]. In order to increase the detection sensitivity, the coupling of NACE with ESI-MS has been investigated and better sensitivity and selectivity were successfully achieved [14, 23]. However, besides the complicated technical knowledge required, the coupling with MS is expensive and not easily accessible especially in Third World countries. In recent years, an alternative to UV detection has been developed for CE, namely contactless conductivity detection (C4 D). This detection method is suitable for all charged analytes regardless of the presence of distinct chromophores or fluorophores. In addition, C4 D has advantages such as simple instrumentation, portability, and cost effectiveness. Recent review articles on this detection method are available [24–27]. Herein, a comprehensive examination of the potential of NACE-C4 D for the analysis of tamoxifen and its major metabolites endoxifen, 4-hydroxytamoxifen, and Ndesmethyltamoxifen is presented. It covers an optimization of the buffer salts and organic solvents as well as a sensitivity comparison between UV and C4 D detection. The validated new approach was demonstrated for the quantification of tamoxifen and its metabolites in human plasma samples collected from breast cancer patients.

2 Materials and methods 2.1 Chemicals and reagents Methanol (MeOH), glacial acetic acid, ethyl acetate, N-methylformamide (NMF), absolute ethanol (EtOH), 2-propanol, ACN, DMSO, and DMF were purchased from Merck (Darmstadt, Germany). MOPS, Tris, CHES, ammonium acetate, tetraethylammonium hydroxide solution, sodium hydroxide, deoxycholic acid sodium salt (NaDCHA), sodium cholate, and 3-(N-morpholino)propanesulfonic acid sodium salt were obtained from Sigma-Aldrich (Buchs, Switzerland). 18-Crown-6 and 4-(dimethylamino)pyridine were purchased from Fluka (Buchs, Switzerland). Tamoxifen, N-desmethyltamoxifen, endoxifen (1:1 E/Z mixture), and (Z)4-hydroxytamoxifen were purchased from Toronto Research Chemicals (Toronto, Canada). Stock solutions of each drug at a concentration of 500 ␮g/mL were prepared in MeOH and kept at 4°C. Working standard solutions at lower concentrations were prepared by dilution in MeOH. The buffer solutions were prepared by mixing appropriate amounts of the respective salts with acetic acid and 18-crown-6.  C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Electrophoresis 2015, 36, 2713–2719

2.2 Apparatus and procedures The CE experiments were performed on a PrinCE 500 series instrument (Prince Technologies, Emmen, The Netherlands) equipped with a C4 D (eDAQ, Denistone East, NSW, Australia) and a Spectra 100 UV–Vis detector (Thermo Separation Products, San Jose, CA, USA). A bare fused-silica capillary of 25 ␮m id and 365 ␮m od with total and effective lengths of 62 and 55 cm, respectively, was employed (Polymicro, Phoenix, AZ, USA). The new capillary was conditioned by flushing with 1 M NaOH for 15 min, water for 5 min, and buffer solution for 10 min. After each analysis run, the capillary was rinsed with the running buffer solution for 5 min to maintain the repeatability of the analysis. The study on the effect of different BGEs was conducted using running buffers consisting of 7.5 mM of diverse buffer salts, 15 mM acetic acid, and 1 mM 18-crown-6 prepared in MeOH. The study on the effect of different organic solvents was performed using 7.5 mM NaDCHA, 15 mM acetic acid, and 1 mM 18-crown-6 as buffer salt dissolved in various organic solvents. The samples were injected hydrodynamically at 150 mbar for 12 s and the separation voltage was set at +25 kV in all studies. Data acquisition was carried out using the software package Chart (eDAQ, NSW, Australia) using a low-pass filter with 2 Hz cutoffs at the input. The S/N ratios were calculated by dividing the peak heights by the SDs of the baseline noise for a 1-min segment (the latter value is provided as a feature of the software package Chart). The HPLC separations [12] were carried out on a Hypersil Gold C18 column (150 × 4.6 mm, particle diameter 5 ␮m), from Thermo Scientific (San Jose, CA, USA) using a Merck Hitachi LaChrom L-7100 pump (Darmstadt, Germany) for mobile phase delivery. A Merck Hitachi L-7200 series autosampler was used for sample introduction. Analyte peaks were detected using a Merck Hitachi L7400 series UV detector and were recorded on a PowerChrome 280 data acquisition system (eDAQ). Separations were carried out using a mobile phase of ACN–5 mM triethylammonium phosphate buffer (pH 3.3) at a ratio of 43:57 v/v. The flow rate was set at 1 mL/min from 0 to 6 min, and increased to 1.3 mL/min from 6.1 min until the end of analysis. UV detection of analytes was at 280 nm.

2.3 Human plasma samples The human plasma samples were collected from five women volunteers participating in a clinical study. All plasma samples were kept at –20°C in a freezer until the experiments were performed. The plasma pretreatment procedures were adopted from reported work with slight modifications [28]. In brief, 1 mL of the plasma sample was placed in a clean tube and 1 mL of 1% formic acid was added. The tube was vortexed for 30 s. The mixture was then extracted with 3 mL of ethyl acetate:2-propanol (95:5, v/v) solvent mixture and was vortexed for 5 min. The mixture was then centrifuged at 5000 × g for 5 min at 4°C. The organic phase was transwww.electrophoresis-journal.com

CE and CEC

Electrophoresis 2015, 36, 2713–2719

2 A

34

5

3 Results and discussion

1 mV

6

2715

3.1 Buffer salt 2

1 B

1

34

2

C

1

5 6

34

5

6

5

2

Electropherograms for the separation of the tamoxifen and its metabolites in five different methanolic buffer salts are shown in Fig. 1. Two separation runs were performed in the newly reported cholate and deoxycholate solution [29, 30], while for two of the measurements more commonly used buffer substances, MOPS with sodium and tetraethylammonium as counter ions, were employed. The most popular salt in NACE, ammonium acetate [20], was also investigated. Note that acetic acid and 18-crowns-6 were also added into the running buffer mainly to keep the pH low as well as to improve the separation resolution of endoxifen racemate. As can be seen in Fig. 1, all buffer salts tested gave good separation of all target analytes. Note that all analytes were detected as positive peaks, while the 4-(dimethylamino)pyridine was detected as negative peak. This is a normal feature of conductivity detection. A close examination of the electropherogram of Fig. 1 reveals variations in peak height and some differences in baseline stability for the different buffer salts. The relevant data are shown in Table 1. All the buffer salts tested produced a good separation of tamoxifen and its metabolites. Low electrophoretic currents of about 1 ␮A were obtained for all of these electrolytes. These values indicate that the buffer salts were of low conductivity. Among all the salts tested, ammonium acetate showed the highest baseline noise, followed by MOPS-TEA, 3-(N-morpholino)propanesulfonic acid sodium salt, sodium cholate, and NaDCHA. A similar trend was also observed for the electrophoretic current recorded for each buffer salt. Ammonium acetate with the highest electrophoretic current at 1.6 ␮A resulted in the highest baseline noise level of 0.1041, while NaDCHA with an electrophoretic current of 0.7 ␮A showed the lowest baseline noise level of 0.0187. The baseline noise for these two buffer salts differed by almost a factor of 6. It therefore appears that increases in electrophoretic currents as low as 0.1 ␮A can lead to significant increases in baseline noise level.

34

6

D

1

2

1

5 3

6

4

E

Time (Sec)

Figure 1. Electropherograms for the separation of tamoxifen and its metabolites in various buffer salts (7.5 mM) in 100% MeOH: (A) sodium cholate, (B) NaDCHA, (C) 3-(Nmorpholino)propanesulfonic acid sodium salt, (D) CH3 COONH4 , (E) MOPS-TEA. Peaks (10 ␮g/mL): (1) 4-(dimethylamino)pyridine (IS), (2) N-desmethyltamoxifen, (3) endoxifen-i, (4) endoxifen-ii, (5) tamoxifen, (6) 4-hydroxytamoxifen. CE conditions: bare fusedsilica capillary of 25 ␮m id and 62 and 56 cm total and effective lengths; applied voltage, +25 kV; hydrodynamic injection at 150 mbar for 12 s.

ferred to a clean tube and was dried with gentle flow of nitrogen and reconstituted with 25 ␮L of MeOH. The final solution was either injected directly into the NACEC4 D system for measurement or quantified using the HPLC method.

Table 1. Peak heights, baseline noise, and S/N ratios for the separation of tamoxifen and metabolites in different buffer salts

Buffer salt Baseline noise Electrophoretic current (␮A)

MOPS-TEA 0.0526 1.3

Compounds analyzed 4-(Dimethylamino)pyridine N-desmethyltamoxifen Endoxifen-i Endoxifen-ii Tamoxifen 4-Hydroxytamoxifen

Peak height (mV) 1.29 1.93 1.57 1.17 2.84 1.88

CH3 COONH4 0.1041 1.6 S/N 24 37 30 22 54 36

Peak height (mV) 1.61 2.59 1.82 1.75 2.02 1.54

NaMOPS 0.0404 1.1 S/N 16 25 18 17 19 15

Peak height (mV) 2.91 1.55 1.18 1.17 2.03 1.60

NaDCHA 0.0187 0.7 S/N 72 38 29 29 50 40

Peak height (mV) 0.51 1.35 1.04 0.93 1.66 1.42

NaCH 0.0222 0.8 S/N 27 72 56 50 89 76

Peak height (mV) 0.46 1.00 0.71 0.64 1.13 0.97

S/N 21 45 32 29 51 44

CE conditions: electrolyte 7.5 mM and 15 mM acetic acid of various buffer salts in 100% MeOH; applied voltage, +25 kV; hydrodynamic injection at 150 mBar for 12 s; 25 ␮m id; sample concentration, 10 ␮g/mL. NaCH: sodium cholate; NaMOPS: 3-(N-morpholino)propanesulfonic acid sodium salt.

 C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.electrophoresis-journal.com

2716

L. Y. Thang et al.

Electrophoresis 2015, 36, 2713–2719

Figure 2. Effect of solvent system on the S/N response of tamoxifen and its metabolites. Buffer salt: 7.5 mM NaDCHA, 15 mM acetic acid, and 1 mM 18-crown-6. CE conditions are the same as for Fig. 1. (Error bars represent SD of results n = 4.)

previous reports [29, 30]. Complete baseline separation of all target analytes including endoxifen racemate was achieved within 6 min, which is to our knowledge the fastest separation ever reported using the NACE method. Based on the overall results obtained, NaDCHA was selected as the buffer salt for the subsequent experiments.

3.2 Solvents

Figure 3. Contrasting (A) UV detection (254 nm) and (B) C4 D for the tamoxifen and its metabolites. Peaks (5 ␮g/mL): (1) 4-(dimethylamino)pyridine (IS), (2) N-desmethyltamoxifen, (3) endoxifen-i, (4) endoxifen-ii, (5) tamoxifen, (6) 4hydroxytamoxifen. Buffer salt: 7.5 mM NaDCHA, 15 mM acetic acid, and 1 mM 18-crown-6 in 100% MeOH. CE conditions are the same as for Fig. 1.

The differences in peak height are also given in the table and vary by a factor of 2. More important is the S/N ratio, here the ratio between the analyte peak height and the baseline noise, which has a direct bearing on the LOD. The highest S/N ratio was obtained for NaDCHA, mainly due to very low baseline noise. This finding is in good accordance with  C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

The use of the amphiprotic solvents EtOH and NMF, as well as the dipolar aprotic DMSO, DMF, and can, as alternatives to MeOH as solvent was investigated. Note that water was not tested due to the insolubility of the target analytes in this medium. It was found that for NMF high electrophoretic currents were observed and the system was unstable. DMSO and DMF resulted in low currents (⬍0.2 ␮A) with no peaks detected after a 20-min run. Unsatisfactory peak shapes and poor sensitivity were observed when EtOH or ACN were employed. The addition of a second solvent to MeOH was also tested. EtOH, ACN, DMF, and NMF were individually added into MeOH at a ratio of 10:90 and 30:70. The use of a MeOH– water mixture was also examined. However, poor peak shapes and band broadening effects were observed, which might be due to aggregation of hydrophobic tamoxifen and its metabolites [23]. It was noted that the addition of a second solvent into MeOH caused significant disruption in the peak resolution and sensitivity, as observed in the increased level of baseline noise. Porras et al. [31] reported that the addition of a second solvent into MeOH could decrease the intermolecular hydrogen bonding between MeOH molecules, resulting

www.electrophoresis-journal.com

CE and CEC

Electrophoresis 2015, 36, 2713–2719

2717

Table 2. Linear range, regression data, LODs, and repeatability test of analytes in human plasma using NACE-C4 D

Linear range (ng/mL)

Analytes

N-desmethyltamoxifen Endoxifen-i Endoxifen-ii Tamoxifen 4-Hydroxytamoxifen

Correlation coefficient

100–5000 100–5000 100–5000 100–5000 100–5000

0.9937 0.9954 0.9925 0.9922 0.9973

LODa) (ng/mL)

30 40 40 30 25

LOQb) (ng/mL)

Repeatabilityc) RSD (%, n = 4)

90 100 100 90 75

Intraday

Interdayd)

2.8 2.4 3.1 2.2 2.3

6.0 6.1 8.8 7.0 8.2

a) Calculated from S/N = 3. b) Calculated from S/N = 10. c) From peak areas for determination of sample spiked at levels of 500 ng/mL of each drug. d) n = 4 each day for 3 days.

3.3 Concurrent UV detection 0.5 mV

A 6

2 34

5

B

1

100

150

200

250

300

350

Time (sec) Figure 4. Electropherograms for the separation of tamoxifen and its metabolites in spiked human plasma after liquid–liquid extraction. Peaks: (1) 4-(dimethylamino)pyridine (IS), (2) Ndesmethyltamoxifen, (3) endoxifen-i, (4) endoxifen-ii, (5) tamoxifen, (6) 4-hydroxytamoxifen. Buffer salt: 7.5 mM NaDCHA, 15 mM acetic acid, and 1 mM 18-crown-6 in 100% MeOH. CE conditions are the same as for Fig. 1.

in a change of solvent mixture properties and interruption of interactions with target analytes. The results for S/N using different mixtures of MeOH with selected solvents are shown in Fig. 2. As can be seen, the addition of a second solvent to MeOH at 10% decreased the S/N values of all target analytes in all cases studied. The peak heights also differed strongly. A further increase of the second solvent to 30% resulted in even lower S/N ratios being obtained, except for the ACN–MeOH mixture, which had a slight rebound in S/N value. On the basis of these results, MeOH was retained as the solvent.  C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Tamoxifen and its metabolites contain aromatic chromophores in their molecular structure and thus are favorable for UV detection. Nevertheless, UV detection is limited to the use of organic solvents with a low UV cutoff wavelength for good detection sensitivity. On the other hand, the detection sensitivity of C4 D in NACE is not restricted by the choice of organic solvent. Hence, in order to highlight the potential advantages of C4 D in this proposed approach a separation of the tamoxifen and its metabolites using simultaneous C4 D and UV detection was carried out. The results are shown in Fig. 3. It can be seen that all compounds are detectable by UV absorption at 254 nm. Other detection wavelengths were also examined such as 200, 220, 230, and 280 nm; however, 254 nm was found to provide the highest S/N ratios for all the targeted compounds. The slight shift in migration time is due to the different position of the two detectors on the capillary. The S/N values for tamoxifen, endoxifen-i, endoxifen-ii, 4-hydroxytamoxifen, and N-desmethyltamoxifen obtained at a concentration of 5 ␮g/mL using UV and C4 D detection were 4, 7, 6, 6, and 7 for UV and 19, 16, 16, 27, and 23 for C4 D, respectively. Surprisingly, the S/N values obtained using C4 D were at least threefold higher than those for UV detection. Note that a capillary with 25 ␮m id was employed in this study, and thus resulted in a shorter detection path length for UV detection, which limits the sensitivity. On the other hand, the detection sensitivity of C4 D has been reported not be affected by the use of capillaries with id as small as 10 ␮m [32]. The results indicate that for high-resolution separations of chromophore compounds that require capillaries with id less than 50 ␮m UV might have lower sensitivity than C4 D. A further advantage of C4 D is, of course, the ability to detect weakly absorbing analytes, which is not possible using any optical detection system.

3.4 Method validation Validation of the developed approached for the analysis of tamoxifen and its metabolites was performed in spiked human www.electrophoresis-journal.com

2718

L. Y. Thang et al.

Electrophoresis 2015, 36, 2713–2719

Table 3. Relative recovery studies of analytes in human plasma using NACE-C4 D

Analyte

Amount spiked (ng/mL)

Amount found (ng/mL)

Recovery (%, n = 4)

Repeatability, RSD (%, n = 4)

N-Desmethyltamoxifen

100.0 250.0 500.0 100.0 250.0 500.0 100.0 250.0 500.0 100.0 250.0 500.0 100.0 250.0 500.0

99.0 248.5 499.0 98.5 249.0 499.0 99.0 249.0 499.0 100.0 249.5 500.0 99.5 249.5 499.0

99.0 99.4 99.8 98.5 99.6 99.8 99.0 99.6 99.8 100.0 99.8 100.0 99.5 99.8 99.8

2.7 2.8 2.7 2.6 2.2 2.3 2.9 3.1 3.0 2.4 2.5 2.2 2.3 2.5 2.4

Endoxifen-i

Endoxifen-ii

Tamoxifen

4-Hydroxytamoxifen

Table 4. Quantitative results for tamoxifen and its metabolites in plasma samples

Plasma sample

P1 P2 P3 P4 P5 Correlation coefficient (r)

N-desmethyltamoxifen (ng/mL)

Endoxifen-i (ng/mL)

Endoxifen-ii (ng/mL)

Endoxifen (ng/mL)

Tamoxifen (ng/mL)

NACE-C4 D

LC–UV

NACE-C4 D

NACE-C4 D

LC–UVa)

NACE-C4 D

LC–UV

NACE-C4 D

LC–UV

2020 ± 110 1630 ± 85 2900 ± 140 1100 ± 50 3540 ± 160 0.9996

2050 ± 110 1600 ± 85 2860 ± 140 1130 ± 65 3480 ± 165

740 ± 35 760 ± 35 320 ± 15 670 ± 30 130 ± 7 09972b)

530 ± 30 980 ± 55 520 ± 30 630 ± 35 120 ± 8

1190 ± 55 2170 ± 100 1520 ± 75 2500 ± 120 1670 ± 80 0.9986

1210 ± 65 2150 ± 110 1500 ± 80 2550 ± 130 1660 ± 85

670 ± 30 390 ± 20 630 ± 30 470 ± 25 350 ± 15 0.9926

700 ± 35 370 ± 20 610 ± 25 480 ± 25 330 ± 20

1200 1680 750 1150 220

± ± ± ± ±

55 80 40 55 15

4-Hydroxytamoxifen (ng/mL)

a) Endoxifen racemate was detected as single peak. b) Calculated by comparing total concentration of racemate obtained by NACE-C4 D with concentration obtained by LC-UV. Errors are SDs (n = 3).

plasma using optimum conditions with 7.5 mM NaDCHA, 15 mM acetic acid, and 1 mM 18-crown-6 in 100% MeOH as running buffer. Calibration curves were acquired for spiked blank plasma samples at eight concentration levels in the range from 50 to 5000 ng/mL. The curves of normalization of the peak areas obtained for tamoxifen and its metabolites with the peak areas for 4-(dimethylamino)pyridine (internal standard) versus analyte concentration (ng/mL) resulted in a good linearity with satisfactory correlation coefficients. The results are summarized in Table 2. Good linearity was achieved with correlation coefficients ranging from 0.9922 to 0.9973 in the concentration range of 100–5000 ng/mL. The overall LODs and LOQs were in the range of 25–40 and 75–100 ng/mL, respectively. The repeatability of the method for analyte peak areas at a spiked concentration of 500 ng/mL was generally acceptable with the intraday (n = 4) and interday (n = 4 each day for 3 days) RSD of 2.2–3.1% and 6.0–8.8%, respectively. The potential of the newly developed approach for the determination of tamoxifen and its metabolites in biological samples was examined by analyzing spiked human plasma. Typical electropherograms of a blank plasma extract and plasma extract containing tamoxifen and its metabolites are shown in Fig. 4. It can be seen that no interfering peaks  C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

were found in the extracted blank plasma. The method recovery was evaluated by determining the concentrations of the analytes spiked at three different concentrations (100, 250, and 500 ng/mL). The concentrations were obtained by comparison with the calibration curves for peak areas obtained with standards that had also been subjected to the extraction procedure. This compensates for the different degrees of extraction efficiency for the different analyte species. The results are summarized in Table 3, are certainly acceptable, and demonstrate that the extraction efficiencies are stable and not affected by the background matrices of the human plasma.

3.5 Analysis of human plasma samples A total of five plasma samples were obtained from breast cancer patients undertaking a single daily dose of 20 mg tamoxifen, who had been undergoing medical treatment for 12 to 60 months. The plasma samples underwent identical pretreatment procedures before being analyzed for tamoxifen and its metabolites either using quantification by the HPLCUV method or by NACE-C4 D. All plasma samples were measured in triplicates. The results are given in Table 4. The www.electrophoresis-journal.com

Electrophoresis 2015, 36, 2713–2719

overall results obtained using NACE-C4 D are comparable to the results obtained employing the chromatographic technique. Note that the endoxifen racemate in the extract was determined as single compound in LC using a C18 stationary phase. The concentrations of tamoxifen and its metabolites present in human plasma vary among patients. This is attributed to variations in individual tamoxifen metabolism and the genotype of the patients. The correlation coefficients, r, for the two pairs of data were determined for tamoxifen, N-desmethyltamoxifen, endoxifen, and 4-hydroxytamoxifen with values found in the range of 0.9926 to 0.9996, indicating an acceptable agreement.

4 Concluding remarks The quantification of tamoxifen and its major metabolites in human plasma samples from breast cancer patients was carried out successfully by NACE-C4 D. The use of the deoxycholic-based electrolyte, which had been previously found to give superior performance in the separation of lipophilic quaternary ammonium cations and free fatty acid anions in NACE-C4 D, was also found to work for tamoxifen and its metabolites. The use of a capillary with smaller id to improve the separation resolution of endoxifen racemate was proven not to affect C4 D sensitivity, unlike in UV detection. This proposed approach opens the new perspective to develop a simpler and affordable portable point-of-care device to monitor the level of tamoxifen and its metabolites in human plasma samples. The authors would like to thank the Universiti Teknologi Malaysia for a UTM Zamalah scholarship (L. Y. T.), Ministry of Science, Technology & Innovation Malaysia (MOSTI, eScience fund no.: 06-01-06-SF1253) for financial support, and Professor Michael Breadmore and Associate Professor Joselito Quirino from the University of Tasmania for scientific discussions. The authors have declared no conflict of interest.

5 References [1] Hoskins, J. M., Carey, L. A., McLeod, H. L., Nat. Rev. Cancer 2009, 9, 576–586. [2] Westbrook, K., Stearns, V., Pharmacol. Ther. 2013, 139, 1–11. [3] Brauch, H., Murdter, T. E., Eichelbaum, M., Schwab, M., Clin. Chem. 2009, 55, 1770–1782. [4] Tchu, S., Lynch, K., Wu, A. B., in: Langman, L. J., Snozek, C. L. H. (Eds.), LC-MS in Drug Analysis, Humana Press, New York 2012, pp. 211–222. [5] Borgna, J.-L., Rochefort, H., Mol. Cell. Endocrinol. 1980, 20, 71–85. [6] Crewe, H. K., Notley, L. M., Wunsch, R. M., Lennard, M. S., Gillam, E. M. J., Drug Metab. Dispos. 2002, 30, 869–874.

 C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

CE and CEC

2719

[7] Lim, Y. C., Desta, Z., Flockhart, D. A., Skaar, T. C., Cancer Chemother. Pharmacol. 2005, 55, 471–478. [8] Johnson, M. D., Zuo, H., Lee, K. H., Trebley, J. P., Rae, J. M., Weatherman, R. V., Desta, Z., Flockhart, D. A., Skaar, T. C., Breast Cancer Res. Treat. 2004, 85, 151–159. [9] Stearns, V., Johnson, M. D., Rae, J. M., Morocho, A., Novielli, A., Bhargava, P., Hayes, D. F., Desta, Z., Flockhart, D. A., J. Natl. Cancer Inst. 2003, 95, 1758–1764. [10] Wu, X. L., Hawse, J. R., Subramaniam, M., Goetz, M. P., Ingle, J. N., Spelsberg, T. C., Cancer Res. 2009, 69, 1722–1727. [11] Arellano, C., Allal, B., Goubaa, A., Roche, H., Chatelut, E., J. Pharm. Biomed. Anal. 2014, 100, 254–261. [12] Antunes, M. V., Rosa, D. D., Viana, T. D. S., Andreolla, H., Fontanive, T. O., Linden, R., J. Pharm. Biomed. Anal. 2013, 76, 13–20. [13] Aranda, E. O., Esteve-Romero, J., Rambla-Alegre, M., Peris-Vicente, J., Bose, D., Talanta 2011, 84, 314–318. [14] Carter, S. J., Li, X. F., Mackey, J. R., Modi, S., Hanson, J., Dovichi, N. J., Electrophoresis 2001, 22, 2730–2736. [15] Peris-Vicente, J., Ochoa-Aranda, E., Bose, D., EsteveRomero, J., Talanta 2015, 131, 535–540. [16] Mihailescu, R., Aboul-Enein, H. Y., Efstatide, M. D., Biomed. Chromatogr. 2000, 14, 180–183. [17] MacCallum, J., Cummings, J., Dixon, J. M., Miller, W. R., Br. J. Cancer 2000, 82, 1629–1635. [18] Peris-Vicente, J., Carda-Broch, S., Esteve-Romero, J., Anal. Sci. 2014, 30, 925–930. [19] Flores, J. R., Nevado, J. J. B., Salcedo, A. M. C., Diaz, M. P. C., Talanta 2005, 65, 155–162. [20] Sanders, J. M., Cunningham, M. L., Electrophoresis 2002, 23, 502–505. [21] Li, X. F., Carter, S. J., Dovichi, N. J., J. Chromatogr. 2000, 895, 81–85. [22] Sanders, J. M., Burka, L. T., Shelby, M. D., Newbold, R. R., Cunningham, M. L., J. Chromatogr. B 1997, 695, 181–185. [23] Lu, W. Z., Poon, G. K., Carmichael, P. L., Cole, R. B., Anal. Chem. 1996, 68, 668–674. [24] Kuban, P., Hauser, P. C., Electrophoresis 2015, 36, 195–211. [25] Kuban, P., Hauser, P. C., Electrophoresis 2013, 34, 55–69. [26] Elbashir, A. A., Aboul-Enein, H. Y., Biomed. Chromatogr. 2014, 28, 1502–1506. [27] Elbashir, A. A., Aboul-Enein, H. Y., Biomed. Chromatogr. 2010, 24, 1038–1044. [28] Wang, H. B., Zhang, B. C., Sun, X. L., Sun, Y. T., Shi, M. Y., J. Chromatogr. Sci. 2014, 52, 400–405. [29] Buglione, L., See, H. H., Hauser, P. C., Electrophoresis 2013, 34, 2072–2077. [30] Buglione, L., See, H. H., Hauser, P. C., Electrophoresis 2013, 34, 317–323. [31] Porras, S. P., Valko, I. E., Jyske, P., Riekkola, M. L., J. Biochem. Biophys. Methods 1999, 38, 89–102. [32] Mai, T. D., Hauser, P. C., Electrophoresis 2013, 34, 1796–1803.

www.electrophoresis-journal.com

Determination of tamoxifen and its metabolites in human plasma by nonaqueous capillary electrophoresis with contactless conductivity detection.

A new approach for the quantification of tamoxifen and its metabolites 4-hydroxytamoxifen, N-desmethyltamoxifen, and 4-hydroxy-N-desmethyltamoxifen (e...
479KB Sizes 5 Downloads 14 Views