Anal. Chem. 2010, 82, 10186–10193
Tamoxifen Metabolite Isomer Separation and Quantification by Liquid Chromatography-Tandem Mass Spectrometry Malgorzata Jaremko,† Yumi Kasai,† Myra F. Barginear,‡ George Raptis,‡ Robert J. Desnick,† and Chunli Yu*,† Department of Genetics and Genomic Sciences and Department of Hematology and Medical Oncology, Mount Sinai School of Medicine, Fifth Avenue at 100th Street, New York, New York 10029, United States Tamoxifen (Tam), the antiestrogen used to treat estrogen receptor-positive breast cancer is a pro-drug that is converted to its major active metabolites, endoxifen and 4-hydroxy-tamoxifen (4-OH-Tam) by various biotransformation enzymes of which cytochrome P450-2D6 (CYP2D6) is key. The usual Tam dose is 20 mg daily; however, the plasma active metabolite concentrations vary due to common genetic variants encoding the biotransformation enzymes and environmental factors (e.g., concomitant drugs) that inhibit these enzymes. Effective treatment depends on adequate Tam conversion to its active isomers. To monitor metabolite plasma levels, a novel liquid chromatography-tandem mass spectrometry (LC-MS/ MS) method was developed to separate and quantitate Tam, N-desmethyl-tamoxifen (ND-Tam), and tamoxifenN-oxide (Tam-N-oxide), and the E, Z, and Z′ isomers of endoxifen and 4-OH-Tam. Known standards were used to identify each metabolite/isomer. Quantitation of these metabolites in plasma was linear from 0.6 to 2000 nM. Intra- and inter-assay reproducibilities were 0.2-8.4% and 0.6-6.3%, respectively. Accuracy determined by spike experiments with known standards was 86-103%. Endoxifen, 4-OH-Tam, and their isomers were stable in fresh frozen plasma for g6 months. This method provides the first sensitive, specific, accurate, and reproducible quantitation of Tam and its metabolite isomers for monitoring Tam-treated breast cancer patients.
Tam (Z-1-(p-dimethylaminoethoxyphenyl)-1,2-diphenyl-1-butene) is a “pro-drug” that is converted to its therapeutically active metabolites, 4-hydroxy-N-desmethyl-tamoxifen (endoxifen) and 4-hydroxy-tamoxifen (4-OH-Tam) by the cytochrome P450 enzyme, CYP2D6, as well as several other P450 enzymes (Figure 1). Over 90% of Tam is primarily demethylated by CYP3A4/5 to N-desmethyl-tamoxifen (ND-Tam) which is then hydroxylated to endoxifen by CYP2D6. A smaller portion of Tam is hydroxylated by CYP2D6 and other CYP enzymes to 4-OH-Tam and then further demethylated by CYP2D6 to endoxifen. Other minor Tam metabolites include tamoxifen-N-oxide (Tam-N-oxide), R-hydroxytamoxifen, and N-didemethyl-tamoxifen.3,4 The hydroxylation of Tam leads to the formation of Z (Zusammen) and E (Entgegen) isomers of 4-OH-Tam and endoxifen. In addition, 4′-hydroxylation of Tam to Z′-4-OH-Tam and Z′-endoxifen occurs at the parabenzene ring (Figure 1).4,5 Cell culture studies have indicated that the antiestrogenic activities of Z-4-OH-Tam and Z-endoxifen were both 30- to 100fold greater than Tam as measured by inhibition of estrogeninduced increases in cell growth, stimulation of plasminogen activity, or induction of progesterone receptor gene expression.4,6-8 The antiestrogenic activities of endoxifen and 4-OH-Tam are approximately equivalent. However, endoxifen is present in the plasma of patients at levels 6- to 12-fold greater than that of 4-OHTam, and unlike 4-OH-Tam, endoxifen binding leads to the degradation of the estrogen receptor.9-11 The E forms of these
Tamoxifen (Tam) therapy has proven to be extremely effective for patients with estrogen receptor-positive breast cancer. Among women treated for 5 years, Tam therapy resulted in a 46% reduction in the recurrence rate and a 28% reduction in the death rate.1,2 Despite its success, many women on Tam therapy relapse and die from progressive disease. Consequently, Tam resistance remains a major clinical problem in breast cancer management.
(3) Goetz, M. P.; Kamal, A.; Ames, M. M. Clin. Pharmacol. Ther. 2008, 83, 160–166. (4) Brauch, H.; Jordan, V. C. Eur. J. Cancer. 2009, 45, 2274–2283. (5) Crewe, H. K.; Notley, L. M.; Wunsch, R. M.; Lennard, M. S.; Gillam, E. M. Drug Metab. Dispos. 2002, 30, 869–874. (6) 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. (7) Robertson, D. W.; Katzenellenbogen, J. A.; Long, D. J.; Rorke, E. A.; Katzenellenbogen, B. S. J. Steroid Biochem. 1982, 16, 1–13. (8) Jordan, V. C. Breast Cancer Res. Treat. 1982, 2, 123–138. (9) Goetz, M. P.; Knox, S. K.; Suman, V. J.; Rae, J. M.; Safgren, S. L.; Ames, M. M.; Visscher, D. W.; Reynolds, C.; Couch, F. J.; Lingle, W. L.; Weinshilboum, R. M.; Fritcher, E. G.; Nibbe, A. M.; Desta, Z.; Nguyen, A.; Flockhart, D. A.; Perez, E. A.; Ingle, J. N. Breast Cancer Res. Treat. 2007, 101, 113–121. (10) Flockhart, D. Clin. Adv. Hematol. Oncol. 2008, 6, 493–494. (11) Wu, X.; Hawse, J. R.; Subramaniam, M.; Goetz, M. P.; Ingle, J. N.; Spelsberg, T. C. Cancer Res. 2009, 69, 1722–1727.
* To whom correspondence should be addressed. Phone: 212-241-6964. Fax: 212-241-1464. E-mail:
[email protected]. † Department of Genetics and Genomic Sciences. ‡ Department of Hematology and Medical Oncology. (1) Early Breast Cancer Trialists′ Collaborative Group. Lancet 1992, 339, 71– 85. (2) Early Breast Cancer Trialists’ Collaborative Group. Lancet 2005, 365, 1687– 1717.
10186
Analytical Chemistry, Vol. 82, No. 24, December 15, 2010
10.1021/ac102337d 2010 American Chemical Society Published on Web 11/18/2010
Figure 1. Schema of Tam phase I biotransformation.
metabolites have no significant antiestrogenic activities,4,6,8,12 and the antiestrogenic activities of the Z′-4-OH-Tam and Z′-endoxifen have not been reported. Recent reports have identified both genetic and environmental factors that alter CYP2D6 activity and affect Tam treatment outcomes.13-16 CYP2D6 genetic variants and certain medications decrease the enzyme’s activity, significantly reducing the conversion of Tam to 4-OH-Tam and endoxifen, reflected by the reduced plasma endoxifen levels.9,10,14 Of note, breast cancer patients with genetic variants that decrease CYP2D6 activity have a significantly increased risk of recurrence and lower event-free survival (HR, 1.33; 95% CI, 1.06-1.68).4,17 However, the CYP2D6 genetic variants do not explain all of the inter-individual differences in the plasma levels of Tam’s active metabolites. It has been suggested that monitoring plasma Tam metabolite levels may indicate which patients are not adequately treated by the standard Tam dose of 20 mg daily. Previous analytical methods to determine the plasma concentrations of Tam and its metabolites include gas chromatography/ mass spectrometry (GC/MS),18,19 high-performance liquid chro(12) Robertson, D. W.; Katzenellenbogen, J. A.; Hayes, J. R.; Katzenellenbogen, B. S. J. Med. Chem. 1982, 25, 167–171. (13) Jin, Y.; Desta, Z.; Stearns, V.; Ward, B.; Ho, H.; Lee, K. H.; Skaar, T.; Storniolo, A. M.; Li, L.; Araba, A.; Blanchard, R.; Nguyen, A.; Ullmer, L.; Hayden, J.; Lemler, S.; Weinshilboum, R. M.; Rae, J. M.; Hayes, D. F.; Flockhart, D. A. J. Natl. Cancer Inst. 2005, 97, 30–39. (14) Borges, S.; Desta, Z.; Li, L.; Skaar, T. C.; Ward, B. A.; Nguyen, A.; Jin, Y.; Storniolo, A. M.; Nikoloff, D. M.; Wu, L.; Hillman, G.; Hayes, D. F.; Stearns, V.; Flockhart, D. A. Clin. Pharmacol. Ther. 2006, 80, 61–74. (15) Desmarais, J. E.; Looper, K. J. J. Clin. Psychiatry 2009, 70, 1688–1697. (16) Kiyotani, K.; Mushiroda, T.; Hosono, N.; Tsunoda, T.; Kubo, M.; Aki, F.; Okazaki, Y.; Hirata, K.; Takatsuka, Y.; Okazaki, M.; Ohsumi, S.; Yamakawa, T.; Sasa, M.; Nakamura, Y.; Zembutsu, H. Pharmacogenet. Genomics 2010, 20, 565–568. (17) Schroth, W.; Goetz, M. P.; Hamann, U.; Fasching, P. A.; Schmidt, M.; Winter, S.; Fritz, P.; Simon, W.; Suman, V. J.; Ames, M. M.; Safgren, S. L.; Kuffel, M. J.; Ulmer, H. U.; Bolander, J.; Strick, R.; Beckmann, M. W.; Koelbl, H.; Weinshilboum, R. M.; Ingle, J. N.; Eichelbaum, M.; Schwab, M.; Brauch, H. JAMA, J. Am. Med. Assoc. 2009, 302, 1429–1436. (18) Daniel, C. P.; Gaskell, S. J.; Bishop, H.; Nicholson, R. I. J. Endocrinol. 1979, 83, 401–408. (19) Murphy, C.; Fotsis, T.; Pantzar, P.; Adlercreutz, H.; Martin, F. J. Steroid Biochem. 1987, 26, 547–555.
matography (HPLC),20 and most recently liquid chromatographytandem mass spectrometry (LC-MS/MS).21-23 Compared to GC/ MS and HPLC methods, LC-MS/MS quantification is superior in sensitivity and specificity, has shorter analysis times, and does not require prior “derivatization” for the assay. However, none of these methods differentiate the E-, Z-, and Z′- isomers of endoxifen and 4-OH-Tam in human plasma. Here, an LC-MS/MS method is described for the simultaneous separation and quantitation of Tam and its E, Z, and Z′ metabolite isomers in plasma. EXPERIMENTAL SECTION Chemicals and Reagents. Standards for calibration and quality control (QC): Z-tamoxifen (Tam) and E/Z-4-hydroxytamoxifen (1:1) (E/Z-4-OH-Tam) were purchased from SigmaAldrich (St. Louis, MO). E/Z-4-hydroxy-N-desmethyl-tamoxifen (1: 1) (E/Z-endoxifen), Z-4-hydroxy-N-desmethyl-tamoxifen (contains up to 10% E isomer) (Z-endoxifen), Z-N-desmethyl-4′-hydroxytamoxifen (Z′-endoxifen), Z-N-desmethyl-tamoxifen hydrochloride (ND-Tam), Z-4-hydroxy-tamoxifen (Z-4-OH-Tam), Z-4′-hydroxytamoxifen (Z′-4-OH-Tam), Z-tamoxifen-N-oxide (Tam-N-oxide) were purchased from Toronto Research Chemicals Inc. (North York, Ontario, Canada). Internal standards: tamoxifen-13C2, 15N (13C2, 15N- Tam) were from Sigma-Aldrich. E/Z N-desmethyl4-hydroxy-tamoxifen-d5 (1:1) (endoxifen-d5), N-desmethyltamoxifen-d5 (ND-Tam-d5), Z-4-hydroxy-tamoxifen-d5 (Z-4-OHTam-d5) were obtained from Toronto Research Chemicals Inc. Other standards for isomer studies: E-R-hydroxy-tamoxifen (ER-OH-Tam), cis-R-hydroxy-tamoxifen (cis-R-OH-Tam), β-hydroxy-tamoxifen (β-OH-Tam), E-3-hydroxy-tamoxifen (E-3OHTam) were from Toronto Research Chemicals Inc. HPLC grade methanol, water, and formic acid were obtained from Fisher (Pittsburgh, PA). Human serum (defibrinated, delipidated, and (20) Berthou, F.; Dreano, Y. J. Chromatogr. 1993, 616, 117–127. (21) Lim, C. K.; Yuan, Z. X.; Jones, R. M.; White, I. N.; Smith, L. L. J. Pharm. Biomed. Anal. 1997, 15, 1335–1342. (22) Gjerde, J.; Kisanga, E. R.; Hauglid, M.; Holm, P. I.; Mellgren, G.; Lien, E. A. J. Chromatogr., A 2005, 1082, 6–14. (23) Teunissen, S. F.; Rosing, H.; Koornstra, R. H.; Linn, S. C.; Schellens, J. H.; Schinkel, A. H.; Beijnen, J. H. J. Chromatogr., B: Anal. Technol. Biomed. Life Sci. 2009, 877, 2519–2529.
Analytical Chemistry, Vol. 82, No. 24, December 15, 2010
10187
double charcoal stripped) for generating calibrators and QCs was from Monobid Inc. (Lake Forest, CA). Preparation of Standards and Internal Standards. Two sets of 1 mM stock solutions were prepared in methanol for calibration and QC. They were either from different vendors or by different weighings of the same standards. The intermediate stock mixture for making calibrators were diluted from stock solution with methanol containing Tam, ND-Tam, E, Z, and Z′-4-OH-Tam, E, Z, and Z′-endoxifen, and Tam-N-oxide. Since plasma E-4-OH-Tam and E-endoxifen were very low, they were not included in the intermediate stock mixture for QC. The internal standard working solution was prepared in methanol containing 10 nM of 13C2, 15NTam, and ND-Tam-d5, 2 nM of 4-OH-Tam-d5, and E/Z-endoxifend5. All stock solutions were stored dried at -70 °C and protected from light. Preparation of Calibrators and QCs. Charcoal filtered human serum was used as the matrix to prepare calibrators and QCs, as it is similar to the human plasma matrix, free of fibrils and endogenous analytes (tamoxifen metabolites) and commercially available. Calibrators were prepared by spiking charcoal filtered human serum with the intermediate calibration stock mixture. In the patient plasma samples, a wide range of metabolite concentrations were measured. In the literature, Tam, ND-Tam, 4-OH-Tam, and endoxifen ranged from 190-420 nM, 280-800 nM, 3-17 nM, and 14-130 nM, respectively.13,24-26 Clinical metabolites concentrations were used to develop the calibration and QC range standards. Six levels of calibrators encompassed the following concentrations: 30-1800 nM for Tam and ND-Tam; 1-60 nM for the 4-OH-Tam isomers; and 2-120 nM for the endoxifen isomers and Tam-N-oxide. QC low (QCL), medium (QCM), and high (QCH) levels were prepared by spiking the intermediate stock mixture with charcoal filtered human serum as shown in Table 2. Also, 3- and 10-fold diluted QCLs (3 × QCL and 10 × QCL) were prepared to serve as surrogates in the analytical runs so as to test sensitivity of the different analytes in each batch. Sample Preparation. The assay was validated for determination of steady-state plasma Tam metabolite concentrations in patients taking the standard Tam dose of 20 mg/day for at least 3 months. A 500 µL aliquot of the internal standard working solution in methanol was added to 100 µL of patient plasma. After vigorous vortexing, the sample was centrifuged for 2 min at 9,300 relative centrifugal force (RCF). The supernatant was transferred to a microfuge filter tube (82031-356, VWR North America, West Chester, PA) and centrifuged for 2 min at 9,300 RCF. A 100 µL aliquot of the final filtrate was transferred to a well of a 96-well plate, and each sample was diluted with 55 µL of deionized water for analysis. Six levels of calibrators, three levels of QCs, and two sensitivity surrogate samples were processed in parallel with each set of patient samples. LC-MS/MS Conditions. An Agilent QQQ 6460 tandem mass spectrometer coupled with the Agilent 1200 series fast resolution LC system was used (Wilmington, Delaware). Tamoxifen metabo(24) Lim, H. S.; Ju Lee, H.; Seok Lee, K.; Sook Lee, E.; Jang, I. J.; Ro, J. J. Clin. Oncol. 2007, 25, 3837–3845. (25) 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. (26) Gjerde, J.; Hauglid, M.; Breilid, H.; Lundgren, S.; Varhaug, J. E.; Kisanga, E. R.; Mellgren, G.; Steen, V. M.; Lien, E. A. Ann. Oncol. 2008, 19, 56–61.
10188
Analytical Chemistry, Vol. 82, No. 24, December 15, 2010
lites were separated on the Zorbax SB-C18 column (2.1 mm × 50 mm, 1.8 µm particle size, Santa Clara, CA) maintained at 55 °C and eluted with a linear gradient at a flow rate of 0.3 mL/min. Mobile phase A consisted of HPLC grade methanol with 0.1% formic acid and mobile phase B was HPLC grade water with 0.1% formic acid. The gradient was increased from 55% of mobile phase A to 70% of A over 3 min, held at 70% for 3.5 min, and then increased to 95% for 2 min. The column was then reconditioned to the initial 55% of mobile phase A for 3 min. The injection volume of each sample was 2 µL. Total run time was 11.6 min from injection to injection. The retention times for tamoxifen metabolites were consistent over time when a cleanup step with a high organic mobile phase was used at the end of the run, followed by a 3 min postrun for re-equilibration. The intrabatch CV% for retention times of Z-4-OH-Tam, Z-endoxifen, Z′-4-OH-Tam, Z′-endoxifen, ND-Tam, Tam, and Tam-N-oxide was between 0.1 and 2.1%. The interbatch CV% for retention times of these metabolites over a period of 5 months during the validation study on the same column with over 1000 injections was between 2.1 and 10.4%. The Agilent QQQ 6460 tandem mass spectrometer was operated in the positive jet stream electrospray ionization (ESI) mode. Nitrogen was used as the nebulizer, turbo (heater) gas, curtain gas, and collision activated dissociation gas. The capillary voltage was set at +4000 V and nozzle voltage was at +1000 V. The ion source gas temperature was 350 °C with a flow rate of 10 L/min. The jetstream sheath gas temperature was 325 °C with a flow rate of 10 L/min. Tamoxifen metabolites were measured by selective reaction monitoring (SRM). Optimal mass spectrometric settings (fragmentor and collision energy) for each transition were chosen automatically using the Mass Hunter Optimizer program (Santa Clara, CA) by a series of flow injections of pure standards under different conditions. These settings for selective compounds (Tam, Z-endoxifen, and Z′-endoxifen) were verified and confirmed manually through a series of flow injections and by comparing the peak area under different conditions. The optimal settings are listed in Table 1. At least one additional SRM transition was monitored for each compound as a qualifier. Characterization of the Hydroxylated Tam Metabolite Isomers. In patients’ plasma, the selective reaction monitoring (SRM) transition of 388/72 for 4-OH-Tam and 374/58 for endoxifen revealed multiple peaks. In order to identify these additional peaks, the following standards were individually injected onto the column: Z′-endoxifen, Z′-4-OH-Tam, E-R-OH-Tam, cis-R-OH-Tam, β-OH-Tam, E-3-OH-Tam. The chromatograms of the individual standards were overlaid on the patients’ chromatograms. Subsequently, the MS2 scan and product ion scan of these standards were performed and the fragmentors and collision energy were optimized. The unique SRM transitions of these isomers were added to our mass spectrometry program. Patient Samples. Subjects taking 20 mg per day of oral tamoxifen were recruited from The Tisch Cancer Institute at The Mount Sinai School of Medicine and Queens Cancer Center at Queens Hospital. Prior to subject recruitment, the Institutional Review Boards of the Mount Sinai Medical Center and Queens Hospital approved the study. Blood samples for the measurement of tamoxifen and its metabolites were collected after a minimum of 90 days of tamoxifen therapy at 20 mg daily.
Table 1. SRM Transitions and Mass Spectrometer Settings for Tam Metabolites and Their Stable Isotope-Labeled Internal Standardsa compound 4-OH-Tam (E, Z, Z′) Z-4-OH-Tam-d5 endoxifen (E, Z, Z′) endoxifen-d5 (E and Z) ND-Tam ND-Tam-d5 Tam 13
C2,15N-Tam
Tam-N-oxide R-OH-Tam
β-OH-Tam
SRM (m/z)
quant ion vs qualifiers
dwell time (ms)
fragmentor (V)
CE (V)
388/72 388/223 393/72 393/228 374/58 374/223 374/152 379/58 379/228 358/58 358/129 363/58 363/129 372/72 372/129 375/75 375/129 388/58 388/72 388/72 388/325 388/179 388/72 388/185
quant ion
50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50
150 150 150 150 150 150 150 150 150 150 150 150 150 150 150 150 150 150 150 150 150 150 150 150
24 20 24 20 20 12 12 20 12 20 24 20 24 20 24 20 24 28 28 24 25 25 24 25
quant ion quant ion quant ion quant ion quant ion quant ion quant ion quant ion quant ion quant ion
a 4-OH-Tam, 4-hydroxy tamoxifen; endoxifen, N-desmethyl-4-OH-tamoxifen; ND-Tam, N-desmethyl tamoxifen; Tam, tamoxifen; OH-Tam, hydroxyl tamoxifen; SRM, selective reaction monitoring.
Assay Validation Studies. The method was validated for linearity, accuracy, precision, recovery, matrix effect, ion suppression, specificity, carry over, and short-term stability according to the CLSI guideline EP10-A3 for validation of bioanalytical methods. Linearity. The linearity study was performed by spiking charcoal-filtered serum with a mixture containing all tested compounds at the following concentrations: 0.6, 1, 1.5, 5.0, 10, 40, 80, 120, 300, 600, 900, 1200, 1500, 1800, 2000 nM and assaying duplicates for each concentration in three separate runs. The lower limit of quantification (LLOQ) and upper limit of quantification (ULOQ) were defined where the percentage bias (% bias) of accuracy and precision were within ±20%. Accuracy and Precision. The recoveries of standards spiked in QCs at low, medium, and high concentrations using QCL, QCM, and QCH samples were documented for accuracy. % Bias e ±20% at the LLOQ and e ±15% at other concentrations were considered acceptable. Precision or reproducibility was evaluated by intra(n ) 10) and inter- (n ) 10) day assays at low, medium, and high concentrations using QCL, QCM, and QCH samples. The coefficient of variance (% CV) e ±20% at the LLOQ and e ±15% at other concentrations were considered acceptable. Recovery, Matrix Effect, and Ion Suppression. Matrix effect and ion suppression were evaluated for all tested metabolites based on standards published by Matuszewski and colleagues.27 Three sets of samples were used. Set A was a neat standard mix solution containing all analytes at the concentration of QCL, QCM, and QCH. Set B was the extracted plasma samples spiked with the standard mix solution at QCL, QCM, and QCH concentrations. Set C was plasma samples spiked with standards at the same concentrations. Matrix effect (ME) % ) peak area B/A × 100 and (27) Matuszewski, B. K.; Constanzer, M. L.; Chavez-Eng, C. M. Anal. Chem. 2003, 75, 3019–3030.
ion suppression (IS) % ) (1 - B/A) × 100. Total efficiency (TE) % ) C/A × 100. Experiments were performed in triplicate. Sample Carryover. Blank plasma samples extracted with methanol were injected after the highest calibration level, after QCH, and at the end of the batch to check carry over in the system. Chromatograms were evaluated for presence of any peaks at the retention time of analytes including the internal standards. Carryover was considered acceptable if there were no peaks, or signal-to-noise ratio (S/N) less than 10, or the quantity was
