2652 Jacek Musijowski ´ Edyta Piorkowska Piotr J. Rudzki Pharmacology Department, Pharmaceutical Research Institute, Warsaw, Poland Received March 3, 2014 Revised June 20, 2014 Accepted July 1, 2014
J. Sep. Sci. 2014, 37, 2652–2658
Research Article
Determination of sunitinib in human plasma using liquid chromatography coupled with mass spectrometry An original method based on liquid chromatography with single quadrupole electrospray ionization mass spectrometry was developed for the determination of sunitinib in human plasma. The quantitation limit of the method at 0.10 ng/mL is comparable to that of tandem mass spectrometry assays. The handling of all solutions containing sunitinib was performed under low-intensity red light to avoid the isomerization of sunitinib and enable quantitation using a single peak. Liquid–liquid extraction with a mixture of n-hexane/isopropanol (90:10 v/v) allowed recoveries at the level of 70%. Measurements were performed using a Zorbax SB-C18 column (3.0 mm × 150 mm, 3.5 m) and isocratic elution with (A) 0.1% aqueous formic acid and (B) acetonitrile/methanol (80:20 v/v) in an A/B ratio of 55:45 at 35⬚C. Under these conditions, sunitinib is eluted at 3.8 min in 6 min of the total run time. The linearity of the calibration curve ranges from 0.10 to 150 ng/mL. The baseline separation of sunitinib and its primary metabolite, N-des-ethyl sunitinib (SU12662), as well as sharp peak shapes, suggest a possibility of extending the applied methodology to the quantitative determination of both compounds. Isotopically labeled sunitinib was used as the internal standard. All required validation tests met the acceptance criteria and proved the method’s reliability and robustness. The method may be conveniently applied to study the pharmacokinetics of sunitinib in humans. Keywords: Bioanalysis / Mass spectrometry / Pharmacokinetics / Sunitinib / Validation DOI 10.1002/jssc.201400231
Additional supporting information may be found in the online version of this article at the publisher’s web-site
1 Introduction Sunitinib is a multitarget tyrosine kinase inhibitor used in the treatment of patients with metastatic renal cell carcinomas and those with metastatic gastrointestinal stromal tumors with progression or intolerance to imatinib [1]. More recently, sunitinib was compared to interferon, as the current standard treatment in a phase 3 trial conducted in the first-line treatment of renal cell carcinomas [2]. The maximum plasma concentration of sunitinib in humans is generally observed between 6 and 10 h after oral administration. The area under the concentration versus time curve (AUC) and the maximum plasma concentration (Cmax ) increase proportionally to the dose within the range of 25–100 mg. The values of Cmax obtained for a Correspondence: Dr. Jacek Musijowski, Pharmacology Department, Pharmaceutical Research Institute, 8 Rydygiera, 01-793 Warsaw, Poland E-mail: [email protected] Fax: +48 22456-38-38
Abbreviations: QC, quality control; SU12662, N-des-ethyl sunitinib; MTBE, tert-Butylmethyl ether C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
50 mg dose are between 19.9–24.0 ng/mL (http://labeling. pfizer.com/ShowLabeling.aspx?id=607) [3]. The terminal half-lives of sunitinib and its main active metabolite, N-desethyl sunitinib (SU12662), range from 40–60 and 80–110 h, respectively. Sunitinib is mainly metabolized by the P450 enzyme CYP3A4 to SU12662 [4] and, as most drugs metabolized by this enzyme, it exhibits wide variability in pharmacokinetics with the coefficient of variation for clearance of approximately 40%. Sunitinib and SU12662 are bound in 95 and 90% to human plasma proteins, respectively [5]. Methods of sunitinib determination based on UV detection [6,7], due to relatively high lower LOQ values, are applicable only to pharmacokinetic studies in humans after multiple administration. Most of the contemporary methods for the determination of sunitinib in human plasma are based on MS/MS detection [8–21]. In spite of the abundance of MS methods to the best of our knowledge no single quadrupole MS detection has been reported so far. Sunitinib in a solution is sensitive to daylight, UV and sodium light, isomerizing to the E isomer [9]. Only a few of the reports advise general protection of sunitinib solutions from light [6,7,9,12,16], however, no details are given on how the plasma samples were protected during the preparation. www.jss-journal.com
J. Sep. Sci. 2014, 37, 2652–2658
Liquid Chromatography
2653
Only one of these reports [7] provides the exact range of light intensity at which the samples were handled. In the cases where limited lighting conditions were applied, a single peak of sunitinib was identified. If two isomers were observed, sunitinib quantitation was based either on the sum of both isomers [10, 14, 17, 19] or one isomer only [15, 18, 20] with the reported variability of isomers ratio in the range of 17– 27% [15]. Two of the methods [11,13], in spite of both isomers being observed, do not present any comments on sunitinib quantitation. One of the reports [8] provides no information on light conditions although only one signal was observed. In order to facilitate sample handling in the quantitation of sunitinib, new conditions are to be developed. In the present work an original bioanalytical method for the determination of sunitinib in human plasma using LC–MS and a LLE is described. Red light conditions allow for the quantitation using the signal of one isomer only. Its range and sensitivity make this method a suitable tool for studying pharmacokinetics of sunitinib in humans after a single dose oral administration.
phase consisted of 0.1% HCOOH (phase A) and ACN/MeOH (80:20 v/v; phase B) mixed in the ratio A/B 55:45 v/v. The flow rate of the mobile phase was 0.35 mL/min and the isocratic elution was used. The sample volume of 20 L was injected onto the column. The column and the autosampler were maintained at 35 ± 2 and 20 ± 5⬚C, respectively. The total chromatographic run time was 6 min with the retention time of sunitinib approximately 3.8 min.
2 Materials and methods
2.4 Preparation of standards and quality control samples
2.1 Instrumentation and reagents The Shimadzu LC–MS system (Duisburg, Germany) consisting of two LC-10ADVP pumps, an autosampler SIL-HTA, a column oven CTO-10A, a degasser DGU-20A3 and a mass spectrometer LCMS-2010 was used. The data processing software was Shimadzu LCMSsolution version 2.05. Black polypropylene, light-proof, 5 mL microcentrifuge tubes were obtained from Agros Technologies (Elgin, IL, USA). Sunitinib and N-des-ethyl sunitinib (SU12662), were synthesized at the Chemistry Department of the Pharmaceutical Research Institute (Warsaw, Poland) [22]. The internal standard, sunitinib-d10 (IS), was obtained from Alsachim (Illkirch Graffenstaden, France). Acetonitrile (ACN; LC–MS grade), methanol (MeOH; LC–MS grade), and n-hexane (HPLC grade) were purchased from POCh (Gliwice, Poland). tert-Butylmethyl ether (MTBE; HPLC grade) was purchased from Sigma-Aldrich (Steinheim, Germany), sodium carbonate p.a. was obtained from Chempur (Piekary Slaskie, Poland). Formic acid (HCOOH) 98–100% p.a. was purchased from Merck (Darmstadt, Germany). 2-Propanol p.a. was purchased from J.T. Baker (Deventer, Netherlands). Purified water from a Milli-Q system (Millipore, Molsheim, France) was used throughout the study.
2.2 Chromatographic conditions The separation was performed on a Zorbax SB-C18 analytical column (150 mm × 3 mm, 3.5 m) purchased from Agilent Technologies (Santa Clara, CA, USA), which was preceded by a guard column SecurityGuard C18 (4 mm × 3 mm) purchased from Phenomenex (Torrance, CA, USA). The mobile C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
2.3 Mass spectrometric conditions The LC–MS was equipped with an ESI operating in the positive ion mode and single-ion monitoring was used as the data acquisition mode. The ions of sunitinib, the IS, and SU12662 were monitored at m/z ratios of 399.20, 409.20, and 371.15, respectively. The probe high voltage and the detector voltage were set at 2.2 and 1.9 kV, respectively. Nitrogen at 3.5 L/min was used as the nebulizer gas. The block temperature was 240⬚C, whereas the curved desolvation line temperature and voltage were 220⬚C and 70 V, respectively.
All standards and plasma samples were prepared in a darkroom under a low-intensity red light (standard 40 W red-tinted light bulb). All standard solutions were kept in amber glass bottles and when not in use or when transferred outside the darkroom the bottles were additionally kept in a light-proof box. The sunitinib, IS, and SU12662 stock solutions were prepared as follows: 13.4 mg of sunitinib L-malate salt (corresponding to 10.0 mg of sunitinib base), 1.0 mg of the IS, and 10.0 mg of SU12662 were dissolved in 10 mL of 0.1% HCOOH/MeOH (50:50, v/v) and stored in amber glass bottles in a light-proof box, in a freezer < 6⬚C. The working solutions were obtained by the appropriate dilutions of stock solutions with 0.1% HCOOH/MeOH (50:50, v/v). The sunitinib standard in plasma (1000 ng/mL) was prepared by spiking blank human plasma with the working sunitinib solution (10 g/mL). All other plasma standards were prepared by the appropriate dilutions. The calibration standards in human plasma were prepared at the following sunitinib concentrations: 0.10, 0.25, 5.00, 25.0, 50.0, 90.0, 125, and 150 ng/mL. The quality control (QC) samples contained sunitinib at the following concentrations: 0.25, 50.0, and 125 ng/mL. An additional standard (250 ng/mL) was prepared for the dilution test. The plasma standards were stored in a freezer below 14⬚C in 5 mL light-proof, polypropylene microcentrifuge tubes. Sodium citrate was used as an anticoagulant. 2.5 Sample preparation An aliquot of human plasma (0.5 mL) was transferred to an 11 mL glass, screw-cap extraction tube. Next, 40 L of www.jss-journal.com
2654
J. Sep. Sci. 2014, 37, 2652–2658
J. Musijowski et al.
the IS (0.5 g/mL) and 0.5 mL of 2 M sodium carbonate were added. The sample was vortex mixed and then 2 mL of n-hexane/2-propanol (90:10, v/v) mixture were added and agitated for 10 min on a vibrax mixer at 1800 rpm. The sample was then centrifuged for 5 min at 3500 rpm and frozen at 70⬚C. The organic layer was then transferred to a glass tube and evaporated under the stream of nitrogen at 45⬚C. The dry residue was dissolved in 250 L of 0.1% HCOOH/MeOH (50:50, v/v) and shaken for 1 min on a vibrax mixer at 1800 rpm. The solution was transferred to an Eppendorf tube and centrifuged for 5 min at 13 000 rpm. The supernatant was transferred to an amber glass autosampler vial.
2.6 Method validation The validation parameters were defined according to the respective regulatory guidances [23, 24]. The study was performed in compliance with the principles of Good Laboratory Practice. The calculations were based on the application of confidence intervals (C.I.) [25]. F-Snedecor test at ␣ = 0.01 was applied to test the hypothesis on the equality of variances. The validation tests included the evaluation of the selectivity, linearity, accuracy, and precision at the lower LOQ and at three QC concentrations: 0.25, 50.0, 125 ng/mL, matrix effects, recovery, the stability of sunitinib in the stock and working standard solutions as well as the short-term, long-term, and freeze–thaw stability of sunitinib in plasma samples and the stability in the extracted samples in the autosampler. Moreover, a possibility of the back-conversion of SU12662 to sunitinib as well as the dilution integrity and carryover were studied. 2.6.1 Selectivity and back-conversion The selectivity was evaluated with blank human plasma samples from six different sources (including heamolyzed and hyperlipidemic plasma). The assessment of the potential back-conversion of SU12662 to sunitinib during sample preparation or mass spectrometric detection was performed by assaying six plasma samples spiked with SU12662 to obtain its plasma concentration of 50.0 ng/mL. 2.6.2 Linearity The linearity was assessed on the basis of statistical calculations performed for six calibration curves. The calculations included the correlation coefficient, the examination of homoscedasticity as a justification for using the weighted regression model and the analysis of the sum of percentage relative errors as the criterion of the weighing factor selection [26]. 2.6.3 Accuracy and precision The inter- and intra-run accuracies and precisions were tested for the QC samples at three concentrations measured in three C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
analytical batches run on separate days. 90% C.I. for mean accuracies and precisions were calculated with the acceptance criteria of 85–115% for the accuracy and ࣘ15% for the precision. 2.6.4 Recovery, matrix effect, and process efficiency The extraction recovery of sunitinib at each QC level was assessed in six different sources of human plasma, including the hyperlipidemic and hemolyzed plasma. The preextraction spiked plasma samples were obtained by spiking blank plasma with both sunitinib and the IS and conducting the standard preparation procedure. The post-extraction spiked plasma samples were obtained by processing blank plasma up to the evaporation step. Then 250 L of respective sunitinib and the IS mixture in 50:50 v/v MeOH/0.1% HCOOH solution were added. The influence of the matrix effects on the method’s reliability was evaluated according to the EMA guidance [23] as well as Matuszewski et al. [27]. The process efficiency, which depends on both the extraction recovery and the absolute matrix effect, was calculated as the ratio of the sunitinib and the IS peak areas in the preextraction spiked plasma samples to that of the respective standard solutions [27]. 2.6.5 Stability The stabilities of sunitinib in plasma at different conditions were assessed for the QCs at two concentrations: 0.25 and 125 ng/mL. Each result was tested for ±15% deviation from the nominal concentration. 90% C.I. for mean stability was calculated with the acceptance criterion of 85–115%. The stabilities of the stock and working solutions of sunitinib and IS were tested with the use of 90% C.I. for mean stability with the acceptance criterion of 90–110%. 2.6.6 Dilution integrity and carryover The dilution integrity test was performed for the 250 ng/mL plasma samples. The sample was diluted 1:1 v/v with control plasma. The measured concentration was then multiplied by 2 and compared to the nominal concentration of 250 ng/mL. The acceptance criteria were 90% C.I. within 85–115% for accuracy and ࣘ15% for precision. The carryover was assessed by sequential injections of the highest calibration standard (150 ng/mL) and blank plasma extracts (n = 6).
3 Results 3.1 Optimization of LC–MS conditions MS parameters were optimized in the positive ion mode with respect to the signal intensity at m/z 399.20, corresponding to sunitinib molecular ion (see Fig. 1). The signals originating from the molecular ions were the most abundant for all the compounds. Fragment ions observed at m/z 283.10 and www.jss-journal.com
Liquid Chromatography
J. Sep. Sci. 2014, 37, 2652–2658
Figure 1. Mass spectrum of sunitinib with two fragment ions also present in the mass spectra of the metabolite and the IS.
326.15 were present in the mass spectra of all three compounds and corresponded to the common fragments of the molecules. Sunitinib and the IS were monitored in all measurements throughout the study. SU12662 (m/z 371.15) was monitored only during the back-conversion test. The initial mobile phase composition was a mixture of pure ACN and 0.1% formic acid in the ratio 60:40, v/v. The flow rate was 0.3 mL/min. Considering that the sunitinib signal (retention time 3.3 min) was observed very close to the baseline disturbance associated with the column dead time, the amount of organic component in the mobile phase was gradually reduced to 45%, until a satisfactory separation was obtained between the two signals. Pure ACN was then replaced with a mixture of ACN/MeOH (80:20, v/v) in order to extend sunitinib retention time without increasing the content of aqueous component in the mobile phase. Preserving the organic solvent at 45% allowed to maintain the sensitivity and increase the flow rate to 0.35 mL/min with still acceptable back pressure. Replacing the formic acid with acetic acid in the aqueous component of the mobile phase did not result in the significant improvement in the peak height or area. The application of 20 mM ammonium formate buffer (pH 4.0) resulted in a narrower peak and longer retention time, however, the signal’s height and area were reduced by half compared to the results obtained with the use of the formic acid. The phenomenon resulted most probably from the lower sunitinib ionization, hence the formic acid solution was selected as the aqueous component. Optimized chromatographic conditions allowed the baseline separation of sunitinib and SU12662 (see Fig. 2).
2655
Figure 2. Example chromatograms obtained during the validation, registered at m/z 371.15 or 399.20 corresponding to the molecular ions of SU12662 and sunitinib, respectively. (A) and (B) were obtained during the back-conversion test. The chromatograms of a blank sample (C) and sunitinib at LLOQ (0.10 ng/mL) (D) and upper limit of quantification (150 ng/mL) (E) are shown for comparison.
MTBE/n-hexane (50:50, v/v), MTBE/n-hexane (1:99, v/v), 2propanol/n-hexane (50:50, v/v), 2-propanol/n-hexane (30:70, v/v), and 2-propanol/n-hexane (10:90, v/v) were tested. Volumes of 1, 2, and 4 mL of the solution and 5, 10, 15, 20 min extraction times were assessed. The optimal results were obtained for 2 mL of the 2-propanol and n-hexane 10:90 v/v mixture during a 10 min extraction and these conditions were applied throughout the study. Three mixtures of 0.1% HCOOH/MeOH in the ratios 75:25, 50:50, and 25:75 v/v were tested for the dissolution of dry residue after evaporation. The results indicated that only 25% content of MeOH was producing significantly lower peak areas, whereas 50 and 75% content yielded similar efficiencies, therefore the solution containing 50% MeOH was used throughout the study.
3.3 Method validation 3.3.1 Selectivity and back-conversion Considering the absence of interferences at the retention times of sunitinib and IS at m/z 399.20 and 409.20, respectively, sufficient selectivity of the method was proven. The measurements for the plasma samples spiked with SU12662 did not indicate any signals at the retention time of sunitinib, implying a lack of metabolite back-conversion under the applied conditions. 3.3.2 Linearity
3.2 Optimization of extraction procedure The extracting solution, its volume as well as the time of the extraction were selected and optimized using the initial procedure which included a back-extraction step, replaced later by the evaporation/dissolution step due to low recoveries. The extracting solutions of MTBE (100%), C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
The calibration curve was linear in the range from 0.10 to 150 ng/mL regarding the sunitinib to the IS peak area ratio versus the nominal concentration of sunitinib. The curve was obtained by the weighted linear regression analysis with the weighing factor w = 1/x2 , where x was the nominal concentration of sunitinib. The values of regression parameters for www.jss-journal.com
2656
J. Sep. Sci. 2014, 37, 2652–2658
J. Musijowski et al.
Table 1. Intrarun and inter-run accuracy and precision of sunitinib determination (n = 6)
Nominal concentration (ng/mL)
Accuracy (%) / Precision (%) Day 1
Day 2
Day 3
Three days
0.10 (LLOQ) 0.25 5.00 50.0 125
106.3 / 10.52 100.0 / 5.43 99.8 / 2.13 97.0 / 1.67 96.8 / 0.89
103.0 / 7.29 109.0 / 5.27 101.9 / 1.24 96.7 / 1.50 98.1 / 2.23
97.5 / 3.96 100.0 / 5.59 112.6 / 2.56 99.9 / 1.68 94.9 / 1.40
102.3 / 7.26 103.0 / 5.43 104.8 / 1.98 97.9 / 1.62 96.6 / 1.51
Table 2. Study of the matrix effects and extraction recoveries for sunitinib and the IS (n = 6)
0.25 (ng/mL)
50.0 (ng/mL)
125 (ng/mL)
SUN IS
Ratio SUN IS
Ratio SUN IS
Ratio
– –
6.94 5.79
0.57 0.57
0.46 0.43
3.3.5 Stability All the results were within the limits of 85–115% for accuracy. For each of the studied concentration levels, the calculated 90% C.I. for the mean stability fell within the acceptance range of 85–115% (see Table 3). Respective tests also confirmed the stability of sunitinib and the IS in both stock and working solutions during the periods of use (data not shown). 3.3.6 Dilution integrity and carryover The mean accuracy obtained for the diluted plasma samples was 109.7% with the precision of 2.61% (n = 6). None of the blank plasma chromatograms, obtained directly after the measurement of 150 ng/mL standard, showed interfering signals at sunitinib or the IS retention times, proving the absence of carryover.
4 Discussion
a)
RSDMF (%) Relative ME b) (%) Absolute ME b) (%)
– –
– –
– –
– –
– –
95.2 92.6 –
92.2 91.9 –
91.7 91.7 –
65.2 66.9 – Recovery (%) 62.1 61.9 – Process b) efficiency (%)
71.8 71.8 – 66.1 66.0 –
70.6 70.4 – 64.7 64.5 –
b)
a) According to EMA [24]. b) According to Matuszewski et al. [27]. MF, matrix factor; ME, matrix effect; SUN, sunitinib.
the curve, described by the equation: y = ax + b were: a = 0.0473 ± 0.0003, b = 0.0014 ± 0.0001, and r = 0.998 (n = 6).
3.3.3 Accuracy and precision The results of the intrarun (within 1 day) and inter-run (within three days) accuracy and precision, expressed as mean values, are presented in Table 1. Additionally, for each concentration level, 90% C.I. fell within the acceptance criteria of 85–115% for the accuracy and ࣘ15% for the precision (data not shown).
3.3.4 Recovery, matrix effect, and process efficiency The comparison of the peak areas determined in the preand post-extraction spiked plasma samples indicated the stability of extraction recovery of sunitinib across the studied concentration range. Moreover, sunitinib concentration did not influence the recovery of the IS (see Table 2). The RSD obtained for normalized matrix factors [23] as well as the relative matrix effect were far below the acceptance criteria of 15% (see Table 2), therefore it was concluded that the matrix effects do not affect the method’s reliability. C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
There were over 214 clinical trials of sunitinib conducted during the last five years, including 20 studies in the year 2013 alone (Clinicaltrials, http://clinicaltrials.gov). Further studies are scheduled and ongoing investigations demonstrate the need for simple and reliable bioanalytical methods for the determination of sunitinib in human plasma. MS/MS detectors due to their selectivity and sensitivity are currently regarded as a standard in pharmacokinetic studies, especially in cases of relatively complex sample matrices like blood serum or plasma [28, 29]. Nonetheless, as illustrated by the present report, single quadrupole MS with appropriate sample preparation are able to detect sub nanogram per milliliter concentrations and are still a valuable tool in bioanalysis. The developed method allows the convenient LC–MS determination of sunitinib in human plasma in the range of 0.10–150 ng/mL. The obtained lower limit of quantification (LLOQ) please add to abbreviations is either better [9–11, 14, 15, 17, 19, 20] or comparable [13, 16, 18] to the previously published LC–MS/MS methods with only one exception [12]. The results confirm that the LC–MS method is sufficient to provide the required performance. The methods using UV detection [6, 7] are still very useful, especially when working with relatively high concentrations, however, they are not applicable to pharmacokinetic studies after single administration of a low dose of sunitinib. Due to its speed, the preferred method of sample preparation for sunitinib MS assays is protein precipitation [9–12, 14, 15, 18–20]. However, the extraction step is a good practice even with MS detectors, as it limits the matrix effects, improves selectivity and sensitivity, extends HPLC column life-time as well as protects the LC system and ion source from the impurities. A total number of over 3000 measurements of plasma samples was performed with the developed method using a single analytical column. No significant changes in the chromatographic separation were observed, which indirectly confirmed the proper sample preparation procedure. www.jss-journal.com
Liquid Chromatography
J. Sep. Sci. 2014, 37, 2652–2658
2657
Table 3. The stability of sunitinib samples expressed as 90% C.I. [25] (n = 6)
Stability test Temperature Time
In autosampler room temperature 24 h
Freeze–thaw ࣘ14⬚C 3 cycles
Short-term room temperature 4h
Long-term ࣘ14⬚C 36 days
0.25 ng/mL 125 ng/mL
86.8–97.0% 104.3–105.4%
103.0–111.0% 97.2–99.1%
91.1–101.6% 98.8–100.5%
94.0–108.0% 94.2–97.7%
a)
a) n = 5 for reference samples.
Inconveniences related to the sunitinib light-induced isomerization predominantly consist in the necessity of either the protection from light or quantitation of the isomers in an equilibrium. The use of yellow light conditions was reported in two methods [9, 18]. Other methods either used the protection from light or no information concerning lighting conditions was provided. This is the first report demonstrating the application of red light in sunitinib determination. The approach proved to be very convenient, allowing further protection against isomerization. Sunitinib standard solution kept for 95 min under red light conditions did not present any significant reduction in the peak area (98.9–108.7%) compared to a freshly prepared solution. Moreover, the use of colorless glass was verified during the method validation, as all the sample preparation steps were performed in standard glass test tubes and the E isomer was either under the LOD or at a negligible level of 0.1% of the Z isomer. This not only allows to avoid quantitation based on two peaks, but also greatly facilitates the solution handling, requiring light protection only when in transport. The presented method was designed for application in a bioequivalence study, therefore only the parent drug, sunitinib, was considered and a validation for the metabolite was not attempted. The primary and pharmacologically active metabolite, N-des-ethyl sunitinib (SU12662), was included only in the selectivity experiments. However, considering a baseline separation of sunitinib and N-des-ethyl sunitinib, sharp peak shapes of both compounds and no interferences from the matrix, the methodology shows a potential for simultaneous quantitative determination of both compounds if necessary. The method was fully validated with respect to regulatory guidelines [23,24] and all parameters met the acceptance criteria. A detailed investigation of the matrix effects [23, 27] as well as the application of confidence intervals for the stability assessment [25] contribute to the increased reliability of results. Therefore, the method is suitable for use in the pharmacokinetic studies performed in humans and meets all requirements set in the regulated bioanalysis.
5 Concluding remarks The developed method allows a convenient single quadrupole LC–ESI-MS determination of sunitinib in human plasma at concentrations within the range of 0.10–150 ng/mL. The C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
method presents high sensitivity, an efficient sample cleanup procedure and expands the scope of instrumental setups applicable to the analysis of the drug. The method was validated with respect to regulatory requirements, making it ready for application in preclinical and clinical analysis in the ongoing clinical trials or bioavailability studies, especially when a single administration of a 50 mg dose is used. The application of red light conditions reduced the isomerization of sunitinib and therefore simplified the sample preparation procedure without the risk of quantitation in the presence of two isomers in equilibrium. This work was supported by the European Union (European Regional Development Fund) under the Innovative Economy Operational Programme 2007–2013 [UDA-POIG.01.03.01-14069/08]. The results of this work were presented in part at the Joint Conference of Polish Mass Spectrometry Society and Ger´ Poland, March 4–7, man Mass Spectrometry Society (Poznan, 2012) and at the VIIIth Multidisciplinary Conference on Drug Research (Rawa Mazowiecka, Poland, May 30–June 1, 2012). The authors gratefully acknowledge Professor Andrzej Kutner for his critical reading of the manuscript and the technical assistance of Magdalena Tro´c and Krystyna Serafin-Byczak. The authors ´ would like to thank Dr. Krzysztof Bankowski as well as teams from the Chemistry Department and R&D Analytical Chemistry Department of the Pharmaceutical Research Institute (Warsaw, Poland) for supplying sunitinib and N-des-ethyl sunitinib reference standards. The authors have declared no conflict of interest.
6 References [1] Motzer, R. J., Michaelson, M. D., Redman, B. G., Hudes, G. R., Wilding, G., Figlin, R. A., Ginsberg, M. S., Kim, S. T., Baum, C. M., DePrimo, S. E., Li, J. Z., Bello, C. L., Theuer, C. P., George, D. J., Rini, B. R., J. Clin. Oncol. 2006, 24, 16–24. [2] Motzer, R. J., Hutson, T. E., Tomczak, P., Michaelson, M. D., Bukowski, R. M., Rixe, O., Oudard, S., Negrier, S., Szczylik, C., Kim, S. T., Chen, I., Bycott P. W., Baum, C. M., Figlin R. A., N. Engl. J. Med. 2007, 356, 115–124. [3] Bello, C. L., Garrett, M., Sherman, L., Smeraglia, J., Ryan, B., Toh, M., Cancer Chemoth. Pharm. 2010, 66, 699–707. [4] Adams, V. R., Leggas, M., Clin. Ther. 2007, 29, 1338–1352.
www.jss-journal.com
2658
J. Musijowski et al.
J. Sep. Sci. 2014, 37, 2652–2658
[17] Couchman, L., Birch, M., Ireland, R., Corrigan, A., Wickramasinghe, S., Josephs, D., Spicer, J., Flanagan, R. J., Anal. Bioanal. Chem. 2012, 403, 1685–1695.
[5] Goodman, V. L., Rock, E. P., Dagher, R., Ramchandani, R. P., Abraham, S., Gobburu, J. V. S., Booth, B. P., Verbois, S. L., Morse, D. E., Liang, C. Y., Chidambaram, N., Jiang, J. X., Tang, S., Mahjoob, K., Justice R., Pazdur R., Clin. Cancer Res. 2007, 13, 1367–1373.
[18] Qiu, F., Bian, W., Li, J., Ge, Z., Biomed. Chrom. 2013, 27, 615–621.
[6] Blanchet, B., Saboureau, C., Benichou, A.S., Billemont, B., Taieb, F., Ropert, S., Dauphin, A., Goldwasser, F., Tod, M., Clin. Chim. Acta. 2009, 404, 134–139.
[19] Lankheet, N. A. G., Hillebrand, M. J. X., Rosing, H., Schellens, J. H. M., Beijnen, J. H., Huitema, A. D. R., Biomed. Chrom. 2013, 27, 466–476.
´ N., Izzedine, H., Milano, [7] Etienne-Grimaldi, M.-C., Renee, G., J. Chrom. B. 2009, 877, 3757–3761.
[20] Andriamanana, I., Gana, I., Duretz, B., Hulin, A., J. Chrom. B. 2013, 926, 83–91.
[8] Minkin, P., Zhao, M., Chen, Z., Ouwerkerk, J., Gelderblom, H., Baker, S. D., J. Chrom. B. 2008, 874, 84–88.
[21] Lankheet, N. A. G., Steeghs, N., Rosing, H., Schellens, J. H. M., Heijnen, J. H., Huitema, A. D. R., Ther. Drug Monit. 2013, 35, 168–176.
[9] de Bruijn, P., Sleijfer, S., Lam, M.-H., Mathijssen, R. H .J., Wiemer, E. A. C., Loos, W. J., J. Pharm. Biomed. Anal. 2010, 51, 934–941.
´ ´ [22] Sidoryk, K., Malinska, M., Bankowski, K., Kubiszewski, M., Łaszcz, M., Bodziachowska-Panfil, M., Kossykowska, M., Giller, T., Kutner, A., Wo´zniak, K., J. Pharm. Sci. 2013, 102, 706–716.
[10] Haouala, A., Zanolari, B., Rochat, B., Montemurro, M., Zaman, K., Duchosal, M. A., Ris, H. B., Leyvraz, S., Widmer, N., Decosterd, L. A., J. Chrom. B. 2009, 877, 1982–1996 [11] Honeywell, R., Yarzadah, K., Giovannetti, E., Losekoot, N., Smit E. F., Walraven M., Lind, J. S. W., Tibaldi, C., Verheul, H. M., Peters, G. J., J. Chrom. B. 2010, 878, 1059– 1068.
[23] Guideline on bioanalytical method validation, EMEA/CHMP/EWP/192217/2009 (2011) European Medicines Agency, Committee for Medicinal Products for Human Use (CHMP).
¨ [12] Rodamer, M., Elsinghorst, P. W., Kinzig, M., Gutschow, ¨ M., Sorgel, F., J. Chrom. B. 2011, 879, 695–706.
[24] Guidance for Industry. Bioanalytical Method Validation. (2001) U.S. Department of Health and Human Services. Food and Drug Administration. Center for Drug Evaluation and Research (CDER). Center for Veterinary Medicine.
[13] Bouchet, S., Chauzit, E., Ducint, D., Castaing, N., CanalRaffin, M., Moore, N., Titier, K., Molimard, M., Clin. Chim. Acta. 2011, 412, 1060–1067.
˜ A. C., J. [26] Almeida, A. M., Castel-Branco, M. M., Falcao, Chrom. B. 2002, 774, 215–222.
[14] Henry, H., Sobhi, H. R., Scheibner, O., Bromirski, M., Nimkar, S. B., Rochat, B., Rapid Commun. Mass Spectrom. 2012, 26, 499–509.
´ A., Acta Pol. Pharm. 2008, 65, 743–747. [25] Rudzki, P. J., Les,
[27] Matuszewski, B. K., Constanzer, M. L., Chavez-Eng, C. M., Anal. Chem. 2003, 75, 3019–3030.
¨ [15] Gotze, L., Hegele, A., Metzelder, S. K., Renz, H., Nockher W. A., Clin. Chim. Acta. 2012, 413, 143–149.
[28] Xie, J., Wang, T., Wang, X., Cheng, X., Dong, H., Wang, Y., Zheng, X., Zhou, L., Xing, J., Dong, Y., J. Sep. Sci. 2014, 37, 1396–1403.
[16] Rais, R., Zhao, M., He, P., Xu, L., Deeken, J. F., Rudek, M. A., Biomed. Chrom. 2012, 26, 1315–1324.
[29] Bae, S. H., Lee, J. K., Cho, D. Y., Bae, S. K., J. Sep. Sci. 2014, 37, 1256–1264.
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.jss-journal.com
J. Sep. Sci. 2014, 37, 2652–2658
Research Article
Determination of sunitinib in human plasma using liquid chromatography coupled with mass spectrometry An original method based on liquid chromatography with single quadrupole electrospray ionization mass spectrometry was developed for the determination of sunitinib in human plasma. The quantitation limit of the method at 0.10 ng/mL is comparable to that of tandem mass spectrometry assays. The handling of all solutions containing sunitinib was performed under low-intensity red light to avoid the isomerization of sunitinib and enable quantitation using a single peak. Liquid–liquid extraction with a mixture of n-hexane/isopropanol (90:10 v/v) allowed recoveries at the level of 70%. Measurements were performed using a Zorbax SB-C18 column (3.0 mm × 150 mm, 3.5 m) and isocratic elution with (A) 0.1% aqueous formic acid and (B) acetonitrile/methanol (80:20 v/v) in an A/B ratio of 55:45 at 35⬚C. Under these conditions, sunitinib is eluted at 3.8 min in 6 min of the total run time. The linearity of the calibration curve ranges from 0.10 to 150 ng/mL. The baseline separation of sunitinib and its primary metabolite, N-des-ethyl sunitinib (SU12662), as well as sharp peak shapes, suggest a possibility of extending the applied methodology to the quantitative determination of both compounds. Isotopically labeled sunitinib was used as the internal standard. All required validation tests met the acceptance criteria and proved the method’s reliability and robustness. The method may be conveniently applied to study the pharmacokinetics of sunitinib in humans. Keywords: Bioanalysis / Mass spectrometry / Pharmacokinetics / Sunitinib / Validation DOI 10.1002/jssc.201400231
Additional supporting information may be found in the online version of this article at the publisher’s web-site
1 Introduction Sunitinib is a multitarget tyrosine kinase inhibitor used in the treatment of patients with metastatic renal cell carcinomas and those with metastatic gastrointestinal stromal tumors with progression or intolerance to imatinib [1]. More recently, sunitinib was compared to interferon, as the current standard treatment in a phase 3 trial conducted in the first-line treatment of renal cell carcinomas [2]. The maximum plasma concentration of sunitinib in humans is generally observed between 6 and 10 h after oral administration. The area under the concentration versus time curve (AUC) and the maximum plasma concentration (Cmax ) increase proportionally to the dose within the range of 25–100 mg. The values of Cmax obtained for a Correspondence: Dr. Jacek Musijowski, Pharmacology Department, Pharmaceutical Research Institute, 8 Rydygiera, 01-793 Warsaw, Poland E-mail: [email protected] Fax: +48 22456-38-38
Abbreviations: QC, quality control; SU12662, N-des-ethyl sunitinib; MTBE, tert-Butylmethyl ether C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
50 mg dose are between 19.9–24.0 ng/mL (http://labeling. pfizer.com/ShowLabeling.aspx?id=607) [3]. The terminal half-lives of sunitinib and its main active metabolite, N-desethyl sunitinib (SU12662), range from 40–60 and 80–110 h, respectively. Sunitinib is mainly metabolized by the P450 enzyme CYP3A4 to SU12662 [4] and, as most drugs metabolized by this enzyme, it exhibits wide variability in pharmacokinetics with the coefficient of variation for clearance of approximately 40%. Sunitinib and SU12662 are bound in 95 and 90% to human plasma proteins, respectively [5]. Methods of sunitinib determination based on UV detection [6,7], due to relatively high lower LOQ values, are applicable only to pharmacokinetic studies in humans after multiple administration. Most of the contemporary methods for the determination of sunitinib in human plasma are based on MS/MS detection [8–21]. In spite of the abundance of MS methods to the best of our knowledge no single quadrupole MS detection has been reported so far. Sunitinib in a solution is sensitive to daylight, UV and sodium light, isomerizing to the E isomer [9]. Only a few of the reports advise general protection of sunitinib solutions from light [6,7,9,12,16], however, no details are given on how the plasma samples were protected during the preparation. www.jss-journal.com
J. Sep. Sci. 2014, 37, 2652–2658
Liquid Chromatography
2653
Only one of these reports [7] provides the exact range of light intensity at which the samples were handled. In the cases where limited lighting conditions were applied, a single peak of sunitinib was identified. If two isomers were observed, sunitinib quantitation was based either on the sum of both isomers [10, 14, 17, 19] or one isomer only [15, 18, 20] with the reported variability of isomers ratio in the range of 17– 27% [15]. Two of the methods [11,13], in spite of both isomers being observed, do not present any comments on sunitinib quantitation. One of the reports [8] provides no information on light conditions although only one signal was observed. In order to facilitate sample handling in the quantitation of sunitinib, new conditions are to be developed. In the present work an original bioanalytical method for the determination of sunitinib in human plasma using LC–MS and a LLE is described. Red light conditions allow for the quantitation using the signal of one isomer only. Its range and sensitivity make this method a suitable tool for studying pharmacokinetics of sunitinib in humans after a single dose oral administration.
phase consisted of 0.1% HCOOH (phase A) and ACN/MeOH (80:20 v/v; phase B) mixed in the ratio A/B 55:45 v/v. The flow rate of the mobile phase was 0.35 mL/min and the isocratic elution was used. The sample volume of 20 L was injected onto the column. The column and the autosampler were maintained at 35 ± 2 and 20 ± 5⬚C, respectively. The total chromatographic run time was 6 min with the retention time of sunitinib approximately 3.8 min.
2 Materials and methods
2.4 Preparation of standards and quality control samples
2.1 Instrumentation and reagents The Shimadzu LC–MS system (Duisburg, Germany) consisting of two LC-10ADVP pumps, an autosampler SIL-HTA, a column oven CTO-10A, a degasser DGU-20A3 and a mass spectrometer LCMS-2010 was used. The data processing software was Shimadzu LCMSsolution version 2.05. Black polypropylene, light-proof, 5 mL microcentrifuge tubes were obtained from Agros Technologies (Elgin, IL, USA). Sunitinib and N-des-ethyl sunitinib (SU12662), were synthesized at the Chemistry Department of the Pharmaceutical Research Institute (Warsaw, Poland) [22]. The internal standard, sunitinib-d10 (IS), was obtained from Alsachim (Illkirch Graffenstaden, France). Acetonitrile (ACN; LC–MS grade), methanol (MeOH; LC–MS grade), and n-hexane (HPLC grade) were purchased from POCh (Gliwice, Poland). tert-Butylmethyl ether (MTBE; HPLC grade) was purchased from Sigma-Aldrich (Steinheim, Germany), sodium carbonate p.a. was obtained from Chempur (Piekary Slaskie, Poland). Formic acid (HCOOH) 98–100% p.a. was purchased from Merck (Darmstadt, Germany). 2-Propanol p.a. was purchased from J.T. Baker (Deventer, Netherlands). Purified water from a Milli-Q system (Millipore, Molsheim, France) was used throughout the study.
2.2 Chromatographic conditions The separation was performed on a Zorbax SB-C18 analytical column (150 mm × 3 mm, 3.5 m) purchased from Agilent Technologies (Santa Clara, CA, USA), which was preceded by a guard column SecurityGuard C18 (4 mm × 3 mm) purchased from Phenomenex (Torrance, CA, USA). The mobile C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
2.3 Mass spectrometric conditions The LC–MS was equipped with an ESI operating in the positive ion mode and single-ion monitoring was used as the data acquisition mode. The ions of sunitinib, the IS, and SU12662 were monitored at m/z ratios of 399.20, 409.20, and 371.15, respectively. The probe high voltage and the detector voltage were set at 2.2 and 1.9 kV, respectively. Nitrogen at 3.5 L/min was used as the nebulizer gas. The block temperature was 240⬚C, whereas the curved desolvation line temperature and voltage were 220⬚C and 70 V, respectively.
All standards and plasma samples were prepared in a darkroom under a low-intensity red light (standard 40 W red-tinted light bulb). All standard solutions were kept in amber glass bottles and when not in use or when transferred outside the darkroom the bottles were additionally kept in a light-proof box. The sunitinib, IS, and SU12662 stock solutions were prepared as follows: 13.4 mg of sunitinib L-malate salt (corresponding to 10.0 mg of sunitinib base), 1.0 mg of the IS, and 10.0 mg of SU12662 were dissolved in 10 mL of 0.1% HCOOH/MeOH (50:50, v/v) and stored in amber glass bottles in a light-proof box, in a freezer < 6⬚C. The working solutions were obtained by the appropriate dilutions of stock solutions with 0.1% HCOOH/MeOH (50:50, v/v). The sunitinib standard in plasma (1000 ng/mL) was prepared by spiking blank human plasma with the working sunitinib solution (10 g/mL). All other plasma standards were prepared by the appropriate dilutions. The calibration standards in human plasma were prepared at the following sunitinib concentrations: 0.10, 0.25, 5.00, 25.0, 50.0, 90.0, 125, and 150 ng/mL. The quality control (QC) samples contained sunitinib at the following concentrations: 0.25, 50.0, and 125 ng/mL. An additional standard (250 ng/mL) was prepared for the dilution test. The plasma standards were stored in a freezer below 14⬚C in 5 mL light-proof, polypropylene microcentrifuge tubes. Sodium citrate was used as an anticoagulant. 2.5 Sample preparation An aliquot of human plasma (0.5 mL) was transferred to an 11 mL glass, screw-cap extraction tube. Next, 40 L of www.jss-journal.com
2654
J. Sep. Sci. 2014, 37, 2652–2658
J. Musijowski et al.
the IS (0.5 g/mL) and 0.5 mL of 2 M sodium carbonate were added. The sample was vortex mixed and then 2 mL of n-hexane/2-propanol (90:10, v/v) mixture were added and agitated for 10 min on a vibrax mixer at 1800 rpm. The sample was then centrifuged for 5 min at 3500 rpm and frozen at 70⬚C. The organic layer was then transferred to a glass tube and evaporated under the stream of nitrogen at 45⬚C. The dry residue was dissolved in 250 L of 0.1% HCOOH/MeOH (50:50, v/v) and shaken for 1 min on a vibrax mixer at 1800 rpm. The solution was transferred to an Eppendorf tube and centrifuged for 5 min at 13 000 rpm. The supernatant was transferred to an amber glass autosampler vial.
2.6 Method validation The validation parameters were defined according to the respective regulatory guidances [23, 24]. The study was performed in compliance with the principles of Good Laboratory Practice. The calculations were based on the application of confidence intervals (C.I.) [25]. F-Snedecor test at ␣ = 0.01 was applied to test the hypothesis on the equality of variances. The validation tests included the evaluation of the selectivity, linearity, accuracy, and precision at the lower LOQ and at three QC concentrations: 0.25, 50.0, 125 ng/mL, matrix effects, recovery, the stability of sunitinib in the stock and working standard solutions as well as the short-term, long-term, and freeze–thaw stability of sunitinib in plasma samples and the stability in the extracted samples in the autosampler. Moreover, a possibility of the back-conversion of SU12662 to sunitinib as well as the dilution integrity and carryover were studied. 2.6.1 Selectivity and back-conversion The selectivity was evaluated with blank human plasma samples from six different sources (including heamolyzed and hyperlipidemic plasma). The assessment of the potential back-conversion of SU12662 to sunitinib during sample preparation or mass spectrometric detection was performed by assaying six plasma samples spiked with SU12662 to obtain its plasma concentration of 50.0 ng/mL. 2.6.2 Linearity The linearity was assessed on the basis of statistical calculations performed for six calibration curves. The calculations included the correlation coefficient, the examination of homoscedasticity as a justification for using the weighted regression model and the analysis of the sum of percentage relative errors as the criterion of the weighing factor selection [26]. 2.6.3 Accuracy and precision The inter- and intra-run accuracies and precisions were tested for the QC samples at three concentrations measured in three C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
analytical batches run on separate days. 90% C.I. for mean accuracies and precisions were calculated with the acceptance criteria of 85–115% for the accuracy and ࣘ15% for the precision. 2.6.4 Recovery, matrix effect, and process efficiency The extraction recovery of sunitinib at each QC level was assessed in six different sources of human plasma, including the hyperlipidemic and hemolyzed plasma. The preextraction spiked plasma samples were obtained by spiking blank plasma with both sunitinib and the IS and conducting the standard preparation procedure. The post-extraction spiked plasma samples were obtained by processing blank plasma up to the evaporation step. Then 250 L of respective sunitinib and the IS mixture in 50:50 v/v MeOH/0.1% HCOOH solution were added. The influence of the matrix effects on the method’s reliability was evaluated according to the EMA guidance [23] as well as Matuszewski et al. [27]. The process efficiency, which depends on both the extraction recovery and the absolute matrix effect, was calculated as the ratio of the sunitinib and the IS peak areas in the preextraction spiked plasma samples to that of the respective standard solutions [27]. 2.6.5 Stability The stabilities of sunitinib in plasma at different conditions were assessed for the QCs at two concentrations: 0.25 and 125 ng/mL. Each result was tested for ±15% deviation from the nominal concentration. 90% C.I. for mean stability was calculated with the acceptance criterion of 85–115%. The stabilities of the stock and working solutions of sunitinib and IS were tested with the use of 90% C.I. for mean stability with the acceptance criterion of 90–110%. 2.6.6 Dilution integrity and carryover The dilution integrity test was performed for the 250 ng/mL plasma samples. The sample was diluted 1:1 v/v with control plasma. The measured concentration was then multiplied by 2 and compared to the nominal concentration of 250 ng/mL. The acceptance criteria were 90% C.I. within 85–115% for accuracy and ࣘ15% for precision. The carryover was assessed by sequential injections of the highest calibration standard (150 ng/mL) and blank plasma extracts (n = 6).
3 Results 3.1 Optimization of LC–MS conditions MS parameters were optimized in the positive ion mode with respect to the signal intensity at m/z 399.20, corresponding to sunitinib molecular ion (see Fig. 1). The signals originating from the molecular ions were the most abundant for all the compounds. Fragment ions observed at m/z 283.10 and www.jss-journal.com
Liquid Chromatography
J. Sep. Sci. 2014, 37, 2652–2658
Figure 1. Mass spectrum of sunitinib with two fragment ions also present in the mass spectra of the metabolite and the IS.
326.15 were present in the mass spectra of all three compounds and corresponded to the common fragments of the molecules. Sunitinib and the IS were monitored in all measurements throughout the study. SU12662 (m/z 371.15) was monitored only during the back-conversion test. The initial mobile phase composition was a mixture of pure ACN and 0.1% formic acid in the ratio 60:40, v/v. The flow rate was 0.3 mL/min. Considering that the sunitinib signal (retention time 3.3 min) was observed very close to the baseline disturbance associated with the column dead time, the amount of organic component in the mobile phase was gradually reduced to 45%, until a satisfactory separation was obtained between the two signals. Pure ACN was then replaced with a mixture of ACN/MeOH (80:20, v/v) in order to extend sunitinib retention time without increasing the content of aqueous component in the mobile phase. Preserving the organic solvent at 45% allowed to maintain the sensitivity and increase the flow rate to 0.35 mL/min with still acceptable back pressure. Replacing the formic acid with acetic acid in the aqueous component of the mobile phase did not result in the significant improvement in the peak height or area. The application of 20 mM ammonium formate buffer (pH 4.0) resulted in a narrower peak and longer retention time, however, the signal’s height and area were reduced by half compared to the results obtained with the use of the formic acid. The phenomenon resulted most probably from the lower sunitinib ionization, hence the formic acid solution was selected as the aqueous component. Optimized chromatographic conditions allowed the baseline separation of sunitinib and SU12662 (see Fig. 2).
2655
Figure 2. Example chromatograms obtained during the validation, registered at m/z 371.15 or 399.20 corresponding to the molecular ions of SU12662 and sunitinib, respectively. (A) and (B) were obtained during the back-conversion test. The chromatograms of a blank sample (C) and sunitinib at LLOQ (0.10 ng/mL) (D) and upper limit of quantification (150 ng/mL) (E) are shown for comparison.
MTBE/n-hexane (50:50, v/v), MTBE/n-hexane (1:99, v/v), 2propanol/n-hexane (50:50, v/v), 2-propanol/n-hexane (30:70, v/v), and 2-propanol/n-hexane (10:90, v/v) were tested. Volumes of 1, 2, and 4 mL of the solution and 5, 10, 15, 20 min extraction times were assessed. The optimal results were obtained for 2 mL of the 2-propanol and n-hexane 10:90 v/v mixture during a 10 min extraction and these conditions were applied throughout the study. Three mixtures of 0.1% HCOOH/MeOH in the ratios 75:25, 50:50, and 25:75 v/v were tested for the dissolution of dry residue after evaporation. The results indicated that only 25% content of MeOH was producing significantly lower peak areas, whereas 50 and 75% content yielded similar efficiencies, therefore the solution containing 50% MeOH was used throughout the study.
3.3 Method validation 3.3.1 Selectivity and back-conversion Considering the absence of interferences at the retention times of sunitinib and IS at m/z 399.20 and 409.20, respectively, sufficient selectivity of the method was proven. The measurements for the plasma samples spiked with SU12662 did not indicate any signals at the retention time of sunitinib, implying a lack of metabolite back-conversion under the applied conditions. 3.3.2 Linearity
3.2 Optimization of extraction procedure The extracting solution, its volume as well as the time of the extraction were selected and optimized using the initial procedure which included a back-extraction step, replaced later by the evaporation/dissolution step due to low recoveries. The extracting solutions of MTBE (100%), C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
The calibration curve was linear in the range from 0.10 to 150 ng/mL regarding the sunitinib to the IS peak area ratio versus the nominal concentration of sunitinib. The curve was obtained by the weighted linear regression analysis with the weighing factor w = 1/x2 , where x was the nominal concentration of sunitinib. The values of regression parameters for www.jss-journal.com
2656
J. Sep. Sci. 2014, 37, 2652–2658
J. Musijowski et al.
Table 1. Intrarun and inter-run accuracy and precision of sunitinib determination (n = 6)
Nominal concentration (ng/mL)
Accuracy (%) / Precision (%) Day 1
Day 2
Day 3
Three days
0.10 (LLOQ) 0.25 5.00 50.0 125
106.3 / 10.52 100.0 / 5.43 99.8 / 2.13 97.0 / 1.67 96.8 / 0.89
103.0 / 7.29 109.0 / 5.27 101.9 / 1.24 96.7 / 1.50 98.1 / 2.23
97.5 / 3.96 100.0 / 5.59 112.6 / 2.56 99.9 / 1.68 94.9 / 1.40
102.3 / 7.26 103.0 / 5.43 104.8 / 1.98 97.9 / 1.62 96.6 / 1.51
Table 2. Study of the matrix effects and extraction recoveries for sunitinib and the IS (n = 6)
0.25 (ng/mL)
50.0 (ng/mL)
125 (ng/mL)
SUN IS
Ratio SUN IS
Ratio SUN IS
Ratio
– –
6.94 5.79
0.57 0.57
0.46 0.43
3.3.5 Stability All the results were within the limits of 85–115% for accuracy. For each of the studied concentration levels, the calculated 90% C.I. for the mean stability fell within the acceptance range of 85–115% (see Table 3). Respective tests also confirmed the stability of sunitinib and the IS in both stock and working solutions during the periods of use (data not shown). 3.3.6 Dilution integrity and carryover The mean accuracy obtained for the diluted plasma samples was 109.7% with the precision of 2.61% (n = 6). None of the blank plasma chromatograms, obtained directly after the measurement of 150 ng/mL standard, showed interfering signals at sunitinib or the IS retention times, proving the absence of carryover.
4 Discussion
a)
RSDMF (%) Relative ME b) (%) Absolute ME b) (%)
– –
– –
– –
– –
– –
95.2 92.6 –
92.2 91.9 –
91.7 91.7 –
65.2 66.9 – Recovery (%) 62.1 61.9 – Process b) efficiency (%)
71.8 71.8 – 66.1 66.0 –
70.6 70.4 – 64.7 64.5 –
b)
a) According to EMA [24]. b) According to Matuszewski et al. [27]. MF, matrix factor; ME, matrix effect; SUN, sunitinib.
the curve, described by the equation: y = ax + b were: a = 0.0473 ± 0.0003, b = 0.0014 ± 0.0001, and r = 0.998 (n = 6).
3.3.3 Accuracy and precision The results of the intrarun (within 1 day) and inter-run (within three days) accuracy and precision, expressed as mean values, are presented in Table 1. Additionally, for each concentration level, 90% C.I. fell within the acceptance criteria of 85–115% for the accuracy and ࣘ15% for the precision (data not shown).
3.3.4 Recovery, matrix effect, and process efficiency The comparison of the peak areas determined in the preand post-extraction spiked plasma samples indicated the stability of extraction recovery of sunitinib across the studied concentration range. Moreover, sunitinib concentration did not influence the recovery of the IS (see Table 2). The RSD obtained for normalized matrix factors [23] as well as the relative matrix effect were far below the acceptance criteria of 15% (see Table 2), therefore it was concluded that the matrix effects do not affect the method’s reliability. C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
There were over 214 clinical trials of sunitinib conducted during the last five years, including 20 studies in the year 2013 alone (Clinicaltrials, http://clinicaltrials.gov). Further studies are scheduled and ongoing investigations demonstrate the need for simple and reliable bioanalytical methods for the determination of sunitinib in human plasma. MS/MS detectors due to their selectivity and sensitivity are currently regarded as a standard in pharmacokinetic studies, especially in cases of relatively complex sample matrices like blood serum or plasma [28, 29]. Nonetheless, as illustrated by the present report, single quadrupole MS with appropriate sample preparation are able to detect sub nanogram per milliliter concentrations and are still a valuable tool in bioanalysis. The developed method allows the convenient LC–MS determination of sunitinib in human plasma in the range of 0.10–150 ng/mL. The obtained lower limit of quantification (LLOQ) please add to abbreviations is either better [9–11, 14, 15, 17, 19, 20] or comparable [13, 16, 18] to the previously published LC–MS/MS methods with only one exception [12]. The results confirm that the LC–MS method is sufficient to provide the required performance. The methods using UV detection [6, 7] are still very useful, especially when working with relatively high concentrations, however, they are not applicable to pharmacokinetic studies after single administration of a low dose of sunitinib. Due to its speed, the preferred method of sample preparation for sunitinib MS assays is protein precipitation [9–12, 14, 15, 18–20]. However, the extraction step is a good practice even with MS detectors, as it limits the matrix effects, improves selectivity and sensitivity, extends HPLC column life-time as well as protects the LC system and ion source from the impurities. A total number of over 3000 measurements of plasma samples was performed with the developed method using a single analytical column. No significant changes in the chromatographic separation were observed, which indirectly confirmed the proper sample preparation procedure. www.jss-journal.com
Liquid Chromatography
J. Sep. Sci. 2014, 37, 2652–2658
2657
Table 3. The stability of sunitinib samples expressed as 90% C.I. [25] (n = 6)
Stability test Temperature Time
In autosampler room temperature 24 h
Freeze–thaw ࣘ14⬚C 3 cycles
Short-term room temperature 4h
Long-term ࣘ14⬚C 36 days
0.25 ng/mL 125 ng/mL
86.8–97.0% 104.3–105.4%
103.0–111.0% 97.2–99.1%
91.1–101.6% 98.8–100.5%
94.0–108.0% 94.2–97.7%
a)
a) n = 5 for reference samples.
Inconveniences related to the sunitinib light-induced isomerization predominantly consist in the necessity of either the protection from light or quantitation of the isomers in an equilibrium. The use of yellow light conditions was reported in two methods [9, 18]. Other methods either used the protection from light or no information concerning lighting conditions was provided. This is the first report demonstrating the application of red light in sunitinib determination. The approach proved to be very convenient, allowing further protection against isomerization. Sunitinib standard solution kept for 95 min under red light conditions did not present any significant reduction in the peak area (98.9–108.7%) compared to a freshly prepared solution. Moreover, the use of colorless glass was verified during the method validation, as all the sample preparation steps were performed in standard glass test tubes and the E isomer was either under the LOD or at a negligible level of 0.1% of the Z isomer. This not only allows to avoid quantitation based on two peaks, but also greatly facilitates the solution handling, requiring light protection only when in transport. The presented method was designed for application in a bioequivalence study, therefore only the parent drug, sunitinib, was considered and a validation for the metabolite was not attempted. The primary and pharmacologically active metabolite, N-des-ethyl sunitinib (SU12662), was included only in the selectivity experiments. However, considering a baseline separation of sunitinib and N-des-ethyl sunitinib, sharp peak shapes of both compounds and no interferences from the matrix, the methodology shows a potential for simultaneous quantitative determination of both compounds if necessary. The method was fully validated with respect to regulatory guidelines [23,24] and all parameters met the acceptance criteria. A detailed investigation of the matrix effects [23, 27] as well as the application of confidence intervals for the stability assessment [25] contribute to the increased reliability of results. Therefore, the method is suitable for use in the pharmacokinetic studies performed in humans and meets all requirements set in the regulated bioanalysis.
5 Concluding remarks The developed method allows a convenient single quadrupole LC–ESI-MS determination of sunitinib in human plasma at concentrations within the range of 0.10–150 ng/mL. The C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
method presents high sensitivity, an efficient sample cleanup procedure and expands the scope of instrumental setups applicable to the analysis of the drug. The method was validated with respect to regulatory requirements, making it ready for application in preclinical and clinical analysis in the ongoing clinical trials or bioavailability studies, especially when a single administration of a 50 mg dose is used. The application of red light conditions reduced the isomerization of sunitinib and therefore simplified the sample preparation procedure without the risk of quantitation in the presence of two isomers in equilibrium. This work was supported by the European Union (European Regional Development Fund) under the Innovative Economy Operational Programme 2007–2013 [UDA-POIG.01.03.01-14069/08]. The results of this work were presented in part at the Joint Conference of Polish Mass Spectrometry Society and Ger´ Poland, March 4–7, man Mass Spectrometry Society (Poznan, 2012) and at the VIIIth Multidisciplinary Conference on Drug Research (Rawa Mazowiecka, Poland, May 30–June 1, 2012). The authors gratefully acknowledge Professor Andrzej Kutner for his critical reading of the manuscript and the technical assistance of Magdalena Tro´c and Krystyna Serafin-Byczak. The authors ´ would like to thank Dr. Krzysztof Bankowski as well as teams from the Chemistry Department and R&D Analytical Chemistry Department of the Pharmaceutical Research Institute (Warsaw, Poland) for supplying sunitinib and N-des-ethyl sunitinib reference standards. The authors have declared no conflict of interest.
6 References [1] Motzer, R. J., Michaelson, M. D., Redman, B. G., Hudes, G. R., Wilding, G., Figlin, R. A., Ginsberg, M. S., Kim, S. T., Baum, C. M., DePrimo, S. E., Li, J. Z., Bello, C. L., Theuer, C. P., George, D. J., Rini, B. R., J. Clin. Oncol. 2006, 24, 16–24. [2] Motzer, R. J., Hutson, T. E., Tomczak, P., Michaelson, M. D., Bukowski, R. M., Rixe, O., Oudard, S., Negrier, S., Szczylik, C., Kim, S. T., Chen, I., Bycott P. W., Baum, C. M., Figlin R. A., N. Engl. J. Med. 2007, 356, 115–124. [3] Bello, C. L., Garrett, M., Sherman, L., Smeraglia, J., Ryan, B., Toh, M., Cancer Chemoth. Pharm. 2010, 66, 699–707. [4] Adams, V. R., Leggas, M., Clin. Ther. 2007, 29, 1338–1352.
www.jss-journal.com
2658
J. Musijowski et al.
J. Sep. Sci. 2014, 37, 2652–2658
[17] Couchman, L., Birch, M., Ireland, R., Corrigan, A., Wickramasinghe, S., Josephs, D., Spicer, J., Flanagan, R. J., Anal. Bioanal. Chem. 2012, 403, 1685–1695.
[5] Goodman, V. L., Rock, E. P., Dagher, R., Ramchandani, R. P., Abraham, S., Gobburu, J. V. S., Booth, B. P., Verbois, S. L., Morse, D. E., Liang, C. Y., Chidambaram, N., Jiang, J. X., Tang, S., Mahjoob, K., Justice R., Pazdur R., Clin. Cancer Res. 2007, 13, 1367–1373.
[18] Qiu, F., Bian, W., Li, J., Ge, Z., Biomed. Chrom. 2013, 27, 615–621.
[6] Blanchet, B., Saboureau, C., Benichou, A.S., Billemont, B., Taieb, F., Ropert, S., Dauphin, A., Goldwasser, F., Tod, M., Clin. Chim. Acta. 2009, 404, 134–139.
[19] Lankheet, N. A. G., Hillebrand, M. J. X., Rosing, H., Schellens, J. H. M., Beijnen, J. H., Huitema, A. D. R., Biomed. Chrom. 2013, 27, 466–476.
´ N., Izzedine, H., Milano, [7] Etienne-Grimaldi, M.-C., Renee, G., J. Chrom. B. 2009, 877, 3757–3761.
[20] Andriamanana, I., Gana, I., Duretz, B., Hulin, A., J. Chrom. B. 2013, 926, 83–91.
[8] Minkin, P., Zhao, M., Chen, Z., Ouwerkerk, J., Gelderblom, H., Baker, S. D., J. Chrom. B. 2008, 874, 84–88.
[21] Lankheet, N. A. G., Steeghs, N., Rosing, H., Schellens, J. H. M., Heijnen, J. H., Huitema, A. D. R., Ther. Drug Monit. 2013, 35, 168–176.
[9] de Bruijn, P., Sleijfer, S., Lam, M.-H., Mathijssen, R. H .J., Wiemer, E. A. C., Loos, W. J., J. Pharm. Biomed. Anal. 2010, 51, 934–941.
´ ´ [22] Sidoryk, K., Malinska, M., Bankowski, K., Kubiszewski, M., Łaszcz, M., Bodziachowska-Panfil, M., Kossykowska, M., Giller, T., Kutner, A., Wo´zniak, K., J. Pharm. Sci. 2013, 102, 706–716.
[10] Haouala, A., Zanolari, B., Rochat, B., Montemurro, M., Zaman, K., Duchosal, M. A., Ris, H. B., Leyvraz, S., Widmer, N., Decosterd, L. A., J. Chrom. B. 2009, 877, 1982–1996 [11] Honeywell, R., Yarzadah, K., Giovannetti, E., Losekoot, N., Smit E. F., Walraven M., Lind, J. S. W., Tibaldi, C., Verheul, H. M., Peters, G. J., J. Chrom. B. 2010, 878, 1059– 1068.
[23] Guideline on bioanalytical method validation, EMEA/CHMP/EWP/192217/2009 (2011) European Medicines Agency, Committee for Medicinal Products for Human Use (CHMP).
¨ [12] Rodamer, M., Elsinghorst, P. W., Kinzig, M., Gutschow, ¨ M., Sorgel, F., J. Chrom. B. 2011, 879, 695–706.
[24] Guidance for Industry. Bioanalytical Method Validation. (2001) U.S. Department of Health and Human Services. Food and Drug Administration. Center for Drug Evaluation and Research (CDER). Center for Veterinary Medicine.
[13] Bouchet, S., Chauzit, E., Ducint, D., Castaing, N., CanalRaffin, M., Moore, N., Titier, K., Molimard, M., Clin. Chim. Acta. 2011, 412, 1060–1067.
˜ A. C., J. [26] Almeida, A. M., Castel-Branco, M. M., Falcao, Chrom. B. 2002, 774, 215–222.
[14] Henry, H., Sobhi, H. R., Scheibner, O., Bromirski, M., Nimkar, S. B., Rochat, B., Rapid Commun. Mass Spectrom. 2012, 26, 499–509.
´ A., Acta Pol. Pharm. 2008, 65, 743–747. [25] Rudzki, P. J., Les,
[27] Matuszewski, B. K., Constanzer, M. L., Chavez-Eng, C. M., Anal. Chem. 2003, 75, 3019–3030.
¨ [15] Gotze, L., Hegele, A., Metzelder, S. K., Renz, H., Nockher W. A., Clin. Chim. Acta. 2012, 413, 143–149.
[28] Xie, J., Wang, T., Wang, X., Cheng, X., Dong, H., Wang, Y., Zheng, X., Zhou, L., Xing, J., Dong, Y., J. Sep. Sci. 2014, 37, 1396–1403.
[16] Rais, R., Zhao, M., He, P., Xu, L., Deeken, J. F., Rudek, M. A., Biomed. Chrom. 2012, 26, 1315–1324.
[29] Bae, S. H., Lee, J. K., Cho, D. Y., Bae, S. K., J. Sep. Sci. 2014, 37, 1256–1264.
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.jss-journal.com
