Journal of Chromatographic Science 2015;53:451– 455 doi:10.1093/chromsci/bmt034 Advance Access publication January 26, 2015
Article
Simultaneous Determination of Cortisol, Cortisone, 6b-Hydroxycortisol and 6b-Hydroxycortisone by HPLC Liyun Zheng1†, Xi Luo1*, Lijun Zhu2, Wenzhao Xie2, Shikun Liu2 and Zeneng Cheng1* 1
Research Institute of Drug Metabolism and Pharmacokinetics, School of Pharmaceutical Sciences, Central South University, 410013 Changsha, Hunan, P.R. China and 2The Third Xiangya Hospital, Central South University, 410013 Changsha, Hunan, P.R. China
*Author to whom correspondence should be addressed. Email: [email protected] †
These authors contributed equally to this work.
Received 26 September 2012; revised 10 December 2012
A specific and sensitive method based on high-performance liquid chromatography with ultraviolet absorbance detection (HPLC –UV) was developed for the simultaneous determination of urinary cortisol (F), cortisone (E), 6b-hydroxycortisol (6b-OHF) and 6b-hydroxycortisone (6b-OHE) using dexamethasone as the internal standard. The method involved solid-phase extraction of the five compounds from urine using Oasis HLB Waters cartridges with an elution solvent of ethyl acetate –diethyl ether (5 mL; 4:1, v/v), followed by 1 mol/L of NaOH (1 mL) and 1.0% acetic acid (1 mL). Separation of the five analytes was achieved within 31 min by using a reversed-phase C18 analytical column (200 3 4.6 mm, 5 mm, Agilent). A UV detector operated at 245 nm was used. According to the method validation, inter-run and intra-run precision was below 9.45% and accuracy ranged from 98.16 to 115.50%. The lower limits of quantitation were 5 ng/mL for four analytes. This is the first HPLC method that can simultaneously determine F, E, 6b-OHF and 6b-OHE in human urine. The assay was applied to research the ratio of (6b-OHF 1 6b-OHE)/(F 1 E) as a non-invasive biomarker for the metabolism of tacrolimus.
Introduction Tacrolimus is a 23-membered macrolide antibiotic produced by Streptomyces tsukubaensis, which is highly lipophilic and insoluble in water (1). It is a potent immunosuppressant that is widely used in kidney transplantation for prophylaxis of organ rejection. It belongs to Biopharmaceutical Classification System II (BCS II), and it is a substrate of both cytochrome P450 3A (CYP3A) and P-glycoprotein (P-gp) (2). Monitoring of tacrolimus blood concentrations is of great importance in the management of kidney transplant recipients because of its narrow therapeutic window and highly variable drug pharmacokinetics (3). However, traditional monitoring methods are costly and inconvenient; additionally, they cannot predict the starting dose (4). As a result, it would be convenient to have a biological marker that can predict the metabolism of tacrolimus, because accurate blood concentrations of tacrolimus can be quickly predicted. The C-6b-oxidation of cortisol (F) and cortisone (E) to 6b-hydroxycortisol (6b-OHF) and 6b-hydroxycortisone (6b-OHE) is catalyzed by CYP3A, and it is proposed that interconversions exist between F and E and 6b-OHF and 6b-OHE (5). Furthermore, a significant inverse correlation was found between (6b-OHF þ 6b-OHE)/(F þ E) and cyclosporin blood concentration/dose ratios in 30 renal transplant recipients (r ¼ –0.807, p , 0.05) (6). Cyclosporin is also a substrate of both
CYP3A and P-gp (7). Therefore, the urinary excretion of (6b-OHF þ 6b-OHE)/(F þ E) is likely to be useful as an endogenous biomarker for the disposition of tacrolimus. The measurements of the simultaneous determination of E and F (8– 10) or F and 6b-OHF (11, 12) have been widely investigated by various techniques. Until now, there have been no methods for the simultaneous determination of E, F, 6b-OHE and 6b-OHF. In the present study, the authors’ laboratory has established a high-performance liquid chromatography– ultraviolet absorbance detection (HPLC –UV) method for the simultaneous determination of endogenous E, F, 6b-OHE and 6b-OHF in human urine with dexamethasone (DM) as the internal standard. The method was applied to the evaluation of (6b-OHF þ 6b-OHE)/(F þ E) as a biomarker for the metabolism of tacrolimus.
Experimental Chemicals and reagents F, E, 6b-OHF, 6b-OHE and DM were purchased from Sigma Chemical Co. (St. Louis, MO). They were of at least 98% purity. Acetonitrile and methanol were of LC grade and purchased from Tedia (Fairfield, OH). All other chemicals were of analytical reagent grade and available from commercial sources (Sinopharm Chemical Reagent Co., Shanghai, China).
Preparation for standards F, E, 6b-OHE and DM were reconstituted as a 1 mg/mL solution in methanol. 6b-OHF was reconstituted as a 0.5 mg/mL solution in methanol. All stock solutions were stored at –208C.
HPLC conditions The analysis was conducted on an 1120 Agilent compact LC with UV absorbance detector (Agilent Technologies, Waldbronn, Germany). Separation was performed on a reversed-phase C18 column (200 4.6 mm, 5 mm, Agilent), monitored by UV absorbance at 245 nm and operated at 1.0 mL/min using the following gradient elution of methanol, acetonitrile and pure water. The stepwise gradient elution program was: 13.5– 20% of acetonitrile, 66.5– 46% of water and 20 –34% of methanol at 0 – 25 min; 13.5% of acetonitrile, 66.5% of water and 20% of methanol at 25.1 –31 min. The temperature of the column was maintained at 308C.
# The Author 2015. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
Sample preparations for HPLC To 2 mL of human urine, 50 mL of the solutions were added, which contained known amounts of DM (4,000 ng/mL) in methanol as the internal standard. The urine samples were applied to Oasis HLB extraction cartridges (Waters Corporation, Milford, MA) that had been pretreated with 3 mL of methanol and 3 mL of distilled water. The cartridges were washed with 3 mL of distilled water and eluted with 5 mL of a solution of ethyl acetate –diethyl ether (4:1, v/v). One milliliter of 1 mol/L NaOH saturated with sodium sulfate was added to the organic extract, which was vortex-mixed for 1 min. After centrifuging, 1 mL of 1.0% acetic acid saturated with sodium sulfate was added to the organic extract, which was vortex-mixed for 1 min. The organic extract was evaporated to dryness at 408C under a stream of nitrogen. A solution (100 mL) of acetonitrile –methanol –water (1:2:7, v/v) was added to the residue. A 30 mL portion of the solution was injected into the LC. The peaks of F, E, 6b-OHF, 6b-OHE and DM were measured. Recovery To 2 mL of solutions containing known amounts of F (10, 40 and 160 ng/mL), E (10, 50 and 200 ng/mL), 6b-OHF (10, 50 and 200 ng/mL) and 6b-OHE (10, 50 and 200 ng/mL) in distilled water was added 50 mL of the solution containing known amounts of DM (4,000 ng/mL) in methanol. Five samples of each concentration were prepared. The samples were conducted through the sample preparation procedure described previously. The recovery values were calculated by comparing the peak areas of these compounds before and after the extraction procedures. Calibration graphs To 2 mL of solutions containing known amounts of E (5, 10, 20, 50, 100, 200 and 400 ng/mL), F (5, 10, 20, 40, 80, 160 and 320 ng/mL), 6b-OHE (5, 10, 20, 50, 100, 200 and 400 ng/mL) and 6b-OHF (5, 10, 20, 50, 100, 200 and 400 ng/mL) in distilled water were added 50 mL of the solution containing known amounts of DM (4,000 ng/mL) in methanol. The samples were conducted through the sample preparation procedure described previously. Peak areas of F, E, 6b-OHF, 6b-OHE and DM were measured. The calibration graphs were obtained by an unweighted least-squares linear fitting of the peak area ratios versus the concentrations of F, E, 6b-OHF and 6b-OHE on each analysis of the standard mixtures.
Accuracy and precision The analytical accuracy and precision were examined by adding known amounts of F (10, 40 and 160 ng/mL), E (10, 50 and 200 ng/mL), 6b-OHF (10, 50 and 200 ng/mL) and 6b-OHE (10, 50 and 200 ng/mL) to urine samples containing known amounts of endogenous F, E, 6b-OHF and 6b-OHE. Five samples of each concentration were conducted through the sample preparation for LC, as described previously. The peak area ratios (E/DM, F/ DM, 6b-OHF/DM and 6b-OHF/DM) were measured.
Clinical sample collection Urine and blood samples were collected from 30 renal transplant recipients (15 female and 15 male, aged 20 to 40 years) at 452 Zheng et al.
8:00 a.m. before they were administered tacrolimus on the day they returned to monitor tacrolimus blood concentrations. Therefore, the protocol did not give the patients any additional pain. The patients participating in the study had taken tacrolimus continuously to produce a steady state blood concentration and were required to have relatively stable liver and kidney functions. The experimental protocol was approved by the ethics committee of the School of Pharmaceutical Sciences, Central South University. All patients gave their written informed consent to participate in the study. Urinary E, F, 6b-OHF and 6b-OHE were measured by the previously described HPLC method. Blood concentrations of tacrolimus were determined by microparticle enzyme immunoassay (MEIA). The urinary ratio of (6b-OHF þ 6b-OHE)/(F þ E) was calculated. Results Chromatography Figure 1A shows the representative chromatograms of standard compounds, Figure 1B shows a morning urine sample and Figure 1C shows the urine sample spiked with 6b-OHF (50 ng/ mL), 6b-OHE (50 ng/mL), E (50 ng/mL), F (40 ng/mL) and DM (50 ng/mL). The analytes, 6b-OHF, 6b-OHE, E, F and DM, were well separated under the previously described chromatographic conditions at retention times of 5.6, 6.7, 18.4, 19, 6 and 24.9 min, respectively. The peak shapes were acceptable and completely resolved one from another. No interfering peaks were observed in the urine samples. Linearity and sensitivity of the assay The peak area ratios of 6b-OHF, 6b-OHE, E and F to DM in human urine was linear with respect to the concentrations of the four analytes. A good correlation was found between the observed peak area ratios (y) and the theoretical concentration (x). Unweighted least-squares regression analysis provided typical regression lines: y ¼ 0.008184x þ 0.016877 (r 2 ¼ 0.99560) for 6b-OHF; y ¼ 0.021729x – 0.004091 (r 2 ¼ 0.99667) for 6b-OHE; y ¼ 0.012444x þ 0.038403 (r 2 ¼ 0.99767) for E; y ¼ 0.011752x þ 0.101080 (r 2 ¼ 0.99250) for F. When a signal-to-noise (S/N) ratio of 10.0 or greater was used as a criterion for a significant response, the lower limit of quantification (LLOQ) was established at 5 ng/mL for the four compounds. Extraction efficiency The extraction recovery values were determined at low, medium and high quality control samples of F, E, 6b-OHF and 6b-OHE. Extraction recovery values were also determined at a fixed concentration of DM (200 ng/mL). Results are shown in Table I.
Precision and accuracy of the assay The results shown in Tables II and III indicate that the assay method is reproducible for replicate analyses of E, F, 6b-OHF and 6b-OHE in human urine within the same day and on different days. The accuracy values for between-batch and within-batch studies at low, medium and high concentrations of E, F, 6b-OHF and 6b-OHE in urine were within acceptable limits (Tables II and III).
Figure 1. LC chromatograms of standard compounds (6b-OHF ¼ 400 ng/mL, 6b-OHE ¼ 400 ng/mL, E ¼ 400 ng/mL, F ¼ 320 ng/mL, DM ¼ 100 ng/mL) (A); morning urine sample (6b-OHF ¼ 113.64 ng/mL, 6b-OHE ¼ 22.92 ng/mL, E ¼ 105.56 ng/mL, F ¼ 70.00 ng/mL) (B); urine sample þ standard compounds þ DM (6b-OHF ¼ 130.59 ng/mL, 6b-OHE ¼ 50.27 ng/mL, E ¼ 122.22 ng/mL, F ¼ 50.84 ng/mL) (C).
Application of the method This method was applied to the urine samples collected from 30 renal transplant recipients; (6b-OHF þ 6b-OHE)/(F þ E), 6bOHF/F and 6b-OHE/E were 1.17 + 0.69, 2.76 + 2.36 and 0.51 + 0.84, respectively. The tacrolimus blood concentration/dose ratio (C/D) was 2.28 + 1.14 ng/mL. Correlations between the three ratios and C/D were examined by Spearman’s rank correlation analysis. All statistic analysis was performed by using SPSS 17.0 statistics software and a p value less than 0.05 was considered significant. A significant correlation was found between (6b-OHF þ 6b-OHE)/(F þ E) and C/D (r ¼ –0.824, p , 0.05).
Weaker correlations were also observed between 6b-OHF/F and C/D (r ¼ –0.511, p , 0.05), and 6b-OHE/E and C/D (r ¼ –0.098, p . 0.05). The values of C/D decreased while urinary (6b-OHF þ 6b-OHE)/(F þ E) increased.
Discussion Acetonitrile is effective to separate 6b-OHF and 6b-OHE, but for F and E, the degree of separation is poor. To obtain good separation and resolution, a solvent system composed of acetonitrile – water–methanol was selected as the mobile phase.
Simultaneous Determination of Cortisol, Cortisone, 6b-Hydroxycortisol and 6b-Hydroxycortisone by HPLC 453
Measured concentration (ng/mL)
Accuracy (%)
RSD (%)
9.85 + 0.32 49.08 + 3.29 200.04 + 2.56
98.53 98.16 100.02
3.28 6.70 1.28
11.54 + 1.09 51.58 + 4.39 204.24 + 2.63
115.47 103.11 102.12
9.45 8.52 1.29
losses of both metabolites, NaOH and acetic acid saturated with sodium sulfate, respectively, were applied to the organic extract. Additionally, to remove insoluble substances and minimize impurity peaks, urine samples were centrifuged at 12,000 rpm for 10 min before pretreatment. Generally speaking, the extraction recovery should be stable at high, medium and low concentrations, but the experiment presented here simultaneously determines four analytes, the properties of four analytes were considered synthetically when choosing an extracting agent. The low extraction recovery of F at a low concentration was probably caused by the limited solubility in the extracting agent, so the extracted quantity is limited. Subsequent washing of the organic extracts with alkaline and acidic solutions can also cause the loss of F; additionally, other urinary endogenous substances may influence the extraction, especially at a low concentration. Although the extraction recovery has certain fluctuations, it meets the standard for the analysis of biological samples. Based on a search of the literature, a few studies were reported on the search for tacrolimus metabolism. Kishino et al. (13) found a significant inverse correlation between 6b-OHF/F and C/D of tacrolimus in two liver transplant recipients (r ¼ 0.658, p , 0.05); however, the limitation of this study was the small number of patients. The results obtained in this study presented a more significant inverse relationship between (6b-OHF þ6b-OHE)/(F þ E) and C/D of tacrolimus (r ¼ –0.824, p , 0.05) in Chinese renal transplant recipients, greater than those between 6b-OHF/F and C/D and 6b-OHE/E and C/D, indicating that urinary (6b-OHF þ 6b-OHE)/(F þ E) may be a more appropriate representation for predicting the disposition of tacrolimus. Although immunochemical techniques are highly sensitive, the four endogenous compounds cannot be separated by this method due to the cross-reactivity of the antibodies, because the four structures are highly similar. Gas chromatography –mass spectrometry (MS) and LC– MS can provide acceptable specificity and sensitivity, but the use of these techniques is generally limited for routine analysis, primarily due to the availability of instruments. This paper described a reliable reversed-phase HPLC assay to simultaneously quantify E, F, 6b-OHF and 6b-OHE in human urine. Distinct advantages include good resolution between the parent drug and its metabolite, good resolution of the analytes from endogenous compounds, accurate assay of large numbers of samples and the requirement of only common instruments.
11.55 + 0.34 50.74 + 3.13 199.39 + 3.32
115.50 101.50 99.70
2.95 6.19 1.66
Acknowledgments
10.61 + 0.64 39.68 + 2.74 161.45 + 15.05
106.09 99.20 100.91
6.02 6.91 9.32
The research was funded by National Natural Science Foundation of China (81072700/H3110) and Doctoral Innovation Projects of Hunan Province (CX2011B055).
Table I Extraction Efficiency of 6b-OHF, 6b-OHE, F, E and DM for Urine Samples* Compound
Concentration (ng/mL)
Recovery (%)
RSD† (%)
6b-OHF
10 50 200 10 50 200 10 40 160 10 50 200 200
62.7 + 0.27 75.0 + 0.83 73.5 + 8.46 69.8 + 0.46 77.8 + 1.44 69.1 + 13.1 65.9 + 0.67 83.9 + 2.16 91.2 + 3.02 63.2 + 0.50 86.5 + 1.81 75.8 + 12.5 91.8 + 5.7
4.4 2.1 5.3 6.7 3.9 9.4 10.8 6.4 2.0 7.7 4.3 8.5 7.1
6b-OHE
F
E
DM
*Note: Data represent mean + standard deviation (SD); n ¼ 5. † Relative standard deviation.
Table II Intra-Day Precision and Accuracy of 6b-OHF, 6b-OHE, E and F in Urine Samples* Added concentration (ng/mL) 6b-OHF 10 50 200 6b-OHE 10 50 200 E 10 50 200 F 10 40 160
Measured concentration (ng/mL)
Accuracy (%)
RSD (%)
10.07 + 0.80 49.08 + 1.95 200.06 + 4.26
100.67 98.16 100.03
7.95 3.96 2.13
10.34 + 0.82 51.58 + 1.50 205.12 + 3.37
103.40 103.11 102.56
7.94 2.92 1.64
11.75 + 0.60 50.75 + 1.83 201.09 + 2.08
117.46 101.50 96.17
5.10 3.61 1.03
9.91 + 0.38 39.67 + 1.25 153.87 + 1.36
99.05 99.20 96.17
3.91 3.14 0.88
*Note: Data represent mean + SD; n ¼ 5.
Table III Inter-Day Precision and Accuracy of 6b-OHF, 6b-OHE, E and F in Urine Samples* Added concentration (ng/mL) 6b-OHF 10 50 200 6b-OHE 10 50 200 E 10 50 200 F 10 40 160
*Note: Data represent mean + SD; n ¼ 5.
Through repeated analysis, it was concluded that all compounds were completely separated when the gradient elution of methanol was 20 –34%. The water solubility of 6b-OHF and 6b-OHE is better than F, E and DM, so in the process of removing impurities, to cut the 454 Zheng et al.
References 1. Hane, K., Fujioka, M., Namiki, Y., Kitagawa, T., Kihara, N., Shimatani, K., et al.; Physico-chemical properties of FK-506; Iyakuhin Kenkyu, (1992); 23: 33–43. 2. Zheng, H., Webber, S., Zeevi, A., Schuetz, E., Zhang, J., Bowman, P., et al.; Tacrolimus dosing in pediatric heart transplant patients is
3.
4.
5.
6.
7.
8.
related to CYP3A5 and MDR1 gene polymorphisms; American Journal of Transplantation, (2003); 3: 477– 483. Iwasaki, K; Metabolism of tacrolimus (FK506) and recent topics in clinical pharmacokinetics; Drug Metabolism and Pharmacokinetics, (2007); 22: 328–335. Kuypers, D.R.; Immunosuppressive drug monitoring—What to use in clinical practice today to improve renal graft outcome; Transplant International, (2005); 18: 140– 150. Abel, S.M., Maggs, J.L., Back, D.J., Park, B.K.; Cortisol metabolism by human liver in vitro—I. Metabolite identification and inter-individual variability; Journal of Steroid Biochemistry and Molecular Biology, (1992); 43: 713–719. Luo, X., Zhu, L.J., Wu, W., Sheng, X.X., Cai, N.F., Liu, S.K., et al.; Simultaneous Determination of 6b-hydroxycortisol and 6b-hydroxycortisone in human urine by LC with UV absorbance detection; Chromatographia, (2009); 70: 1215–1219. Jin, M.J., Shimada, T., Yokogawa, K.; Contributions of intestinal P-glycoprotein and CYP3A to oral bioavailability of cyclosporin A in mice treated with or without dexamethasone; International Journal of Pharmaceutics, (2006); 309: 81– 86. Perogamvros, L., Owen, L.J., Newell-Price, J., Ray, D.W., Trainer, P.J., Keevil, B.G.; Simultaneous measurement of cortisol and cortisone in human saliva using liquid chromatography–tandem mass spectrometry: Application in basal and stimulated conditions; Journal of Chromatography B, (2009); 877: 3771–3775.
9. Santos-Montes, A., Gonzalo-Lumbreras, R., Izquierdo-Hornillos, R.; Simultaneous determination of cortisol and cortisone in urine by reversed-phase high-performance liquid chromatography clinical and doping control applications; Journal of Chromatography B, (1995); 673: 27 –33. 10. Goto, J., Shamsa, F., Goto, N., Nambara, T.; The simultaneous determination of serum cortisol and cortisone by highperformance liquid chromatography with fluorimetric detection; Journal of Pharmaceutical and Biomedical Analysis, (1983); 1: 83– 88. 11. Hu, Z.Y., Gong, Q., Hu, X.L., Wang, L.Q., Cao, Y.J., Cao, W.; Simultaneous determination of 6b-hydroxycortisol and cortisol in human, urine and plasma by liquid chromatography with ultraviolet absorbance detection for phenotyping the CYP3A activity; Journal of Chromatography B, (2005); 826: 238–243. 12. Lutz, U., Bittner, N., Ufer, M., Lutz, W.K.; Quantification of cortisol and 6 beta-hydroxycortisol in human urine by LC-MS/MS, and gender-specific evaluation of the metabolic ratio as biomarker of CYP3A activity; Journal of Chromatography B, (2010); 878: 97– 10. 13. Kishino, S., Ogawa, M., Akekuma, Y., Sugawara, M., Shimamura, T., Furukawa, H.; The variability of liver graft function and urinary 6b-hydroxycortisol to cortisol ratio during liver regeneration in liver transplant recipients; Clinical Transplantation, (2004); 18: 124– 129.
Simultaneous Determination of Cortisol, Cortisone, 6b-Hydroxycortisol and 6b-Hydroxycortisone by HPLC 455
Article
Simultaneous Determination of Cortisol, Cortisone, 6b-Hydroxycortisol and 6b-Hydroxycortisone by HPLC Liyun Zheng1†, Xi Luo1*, Lijun Zhu2, Wenzhao Xie2, Shikun Liu2 and Zeneng Cheng1* 1
Research Institute of Drug Metabolism and Pharmacokinetics, School of Pharmaceutical Sciences, Central South University, 410013 Changsha, Hunan, P.R. China and 2The Third Xiangya Hospital, Central South University, 410013 Changsha, Hunan, P.R. China
*Author to whom correspondence should be addressed. Email: [email protected] †
These authors contributed equally to this work.
Received 26 September 2012; revised 10 December 2012
A specific and sensitive method based on high-performance liquid chromatography with ultraviolet absorbance detection (HPLC –UV) was developed for the simultaneous determination of urinary cortisol (F), cortisone (E), 6b-hydroxycortisol (6b-OHF) and 6b-hydroxycortisone (6b-OHE) using dexamethasone as the internal standard. The method involved solid-phase extraction of the five compounds from urine using Oasis HLB Waters cartridges with an elution solvent of ethyl acetate –diethyl ether (5 mL; 4:1, v/v), followed by 1 mol/L of NaOH (1 mL) and 1.0% acetic acid (1 mL). Separation of the five analytes was achieved within 31 min by using a reversed-phase C18 analytical column (200 3 4.6 mm, 5 mm, Agilent). A UV detector operated at 245 nm was used. According to the method validation, inter-run and intra-run precision was below 9.45% and accuracy ranged from 98.16 to 115.50%. The lower limits of quantitation were 5 ng/mL for four analytes. This is the first HPLC method that can simultaneously determine F, E, 6b-OHF and 6b-OHE in human urine. The assay was applied to research the ratio of (6b-OHF 1 6b-OHE)/(F 1 E) as a non-invasive biomarker for the metabolism of tacrolimus.
Introduction Tacrolimus is a 23-membered macrolide antibiotic produced by Streptomyces tsukubaensis, which is highly lipophilic and insoluble in water (1). It is a potent immunosuppressant that is widely used in kidney transplantation for prophylaxis of organ rejection. It belongs to Biopharmaceutical Classification System II (BCS II), and it is a substrate of both cytochrome P450 3A (CYP3A) and P-glycoprotein (P-gp) (2). Monitoring of tacrolimus blood concentrations is of great importance in the management of kidney transplant recipients because of its narrow therapeutic window and highly variable drug pharmacokinetics (3). However, traditional monitoring methods are costly and inconvenient; additionally, they cannot predict the starting dose (4). As a result, it would be convenient to have a biological marker that can predict the metabolism of tacrolimus, because accurate blood concentrations of tacrolimus can be quickly predicted. The C-6b-oxidation of cortisol (F) and cortisone (E) to 6b-hydroxycortisol (6b-OHF) and 6b-hydroxycortisone (6b-OHE) is catalyzed by CYP3A, and it is proposed that interconversions exist between F and E and 6b-OHF and 6b-OHE (5). Furthermore, a significant inverse correlation was found between (6b-OHF þ 6b-OHE)/(F þ E) and cyclosporin blood concentration/dose ratios in 30 renal transplant recipients (r ¼ –0.807, p , 0.05) (6). Cyclosporin is also a substrate of both
CYP3A and P-gp (7). Therefore, the urinary excretion of (6b-OHF þ 6b-OHE)/(F þ E) is likely to be useful as an endogenous biomarker for the disposition of tacrolimus. The measurements of the simultaneous determination of E and F (8– 10) or F and 6b-OHF (11, 12) have been widely investigated by various techniques. Until now, there have been no methods for the simultaneous determination of E, F, 6b-OHE and 6b-OHF. In the present study, the authors’ laboratory has established a high-performance liquid chromatography– ultraviolet absorbance detection (HPLC –UV) method for the simultaneous determination of endogenous E, F, 6b-OHE and 6b-OHF in human urine with dexamethasone (DM) as the internal standard. The method was applied to the evaluation of (6b-OHF þ 6b-OHE)/(F þ E) as a biomarker for the metabolism of tacrolimus.
Experimental Chemicals and reagents F, E, 6b-OHF, 6b-OHE and DM were purchased from Sigma Chemical Co. (St. Louis, MO). They were of at least 98% purity. Acetonitrile and methanol were of LC grade and purchased from Tedia (Fairfield, OH). All other chemicals were of analytical reagent grade and available from commercial sources (Sinopharm Chemical Reagent Co., Shanghai, China).
Preparation for standards F, E, 6b-OHE and DM were reconstituted as a 1 mg/mL solution in methanol. 6b-OHF was reconstituted as a 0.5 mg/mL solution in methanol. All stock solutions were stored at –208C.
HPLC conditions The analysis was conducted on an 1120 Agilent compact LC with UV absorbance detector (Agilent Technologies, Waldbronn, Germany). Separation was performed on a reversed-phase C18 column (200 4.6 mm, 5 mm, Agilent), monitored by UV absorbance at 245 nm and operated at 1.0 mL/min using the following gradient elution of methanol, acetonitrile and pure water. The stepwise gradient elution program was: 13.5– 20% of acetonitrile, 66.5– 46% of water and 20 –34% of methanol at 0 – 25 min; 13.5% of acetonitrile, 66.5% of water and 20% of methanol at 25.1 –31 min. The temperature of the column was maintained at 308C.
# The Author 2015. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
Sample preparations for HPLC To 2 mL of human urine, 50 mL of the solutions were added, which contained known amounts of DM (4,000 ng/mL) in methanol as the internal standard. The urine samples were applied to Oasis HLB extraction cartridges (Waters Corporation, Milford, MA) that had been pretreated with 3 mL of methanol and 3 mL of distilled water. The cartridges were washed with 3 mL of distilled water and eluted with 5 mL of a solution of ethyl acetate –diethyl ether (4:1, v/v). One milliliter of 1 mol/L NaOH saturated with sodium sulfate was added to the organic extract, which was vortex-mixed for 1 min. After centrifuging, 1 mL of 1.0% acetic acid saturated with sodium sulfate was added to the organic extract, which was vortex-mixed for 1 min. The organic extract was evaporated to dryness at 408C under a stream of nitrogen. A solution (100 mL) of acetonitrile –methanol –water (1:2:7, v/v) was added to the residue. A 30 mL portion of the solution was injected into the LC. The peaks of F, E, 6b-OHF, 6b-OHE and DM were measured. Recovery To 2 mL of solutions containing known amounts of F (10, 40 and 160 ng/mL), E (10, 50 and 200 ng/mL), 6b-OHF (10, 50 and 200 ng/mL) and 6b-OHE (10, 50 and 200 ng/mL) in distilled water was added 50 mL of the solution containing known amounts of DM (4,000 ng/mL) in methanol. Five samples of each concentration were prepared. The samples were conducted through the sample preparation procedure described previously. The recovery values were calculated by comparing the peak areas of these compounds before and after the extraction procedures. Calibration graphs To 2 mL of solutions containing known amounts of E (5, 10, 20, 50, 100, 200 and 400 ng/mL), F (5, 10, 20, 40, 80, 160 and 320 ng/mL), 6b-OHE (5, 10, 20, 50, 100, 200 and 400 ng/mL) and 6b-OHF (5, 10, 20, 50, 100, 200 and 400 ng/mL) in distilled water were added 50 mL of the solution containing known amounts of DM (4,000 ng/mL) in methanol. The samples were conducted through the sample preparation procedure described previously. Peak areas of F, E, 6b-OHF, 6b-OHE and DM were measured. The calibration graphs were obtained by an unweighted least-squares linear fitting of the peak area ratios versus the concentrations of F, E, 6b-OHF and 6b-OHE on each analysis of the standard mixtures.
Accuracy and precision The analytical accuracy and precision were examined by adding known amounts of F (10, 40 and 160 ng/mL), E (10, 50 and 200 ng/mL), 6b-OHF (10, 50 and 200 ng/mL) and 6b-OHE (10, 50 and 200 ng/mL) to urine samples containing known amounts of endogenous F, E, 6b-OHF and 6b-OHE. Five samples of each concentration were conducted through the sample preparation for LC, as described previously. The peak area ratios (E/DM, F/ DM, 6b-OHF/DM and 6b-OHF/DM) were measured.
Clinical sample collection Urine and blood samples were collected from 30 renal transplant recipients (15 female and 15 male, aged 20 to 40 years) at 452 Zheng et al.
8:00 a.m. before they were administered tacrolimus on the day they returned to monitor tacrolimus blood concentrations. Therefore, the protocol did not give the patients any additional pain. The patients participating in the study had taken tacrolimus continuously to produce a steady state blood concentration and were required to have relatively stable liver and kidney functions. The experimental protocol was approved by the ethics committee of the School of Pharmaceutical Sciences, Central South University. All patients gave their written informed consent to participate in the study. Urinary E, F, 6b-OHF and 6b-OHE were measured by the previously described HPLC method. Blood concentrations of tacrolimus were determined by microparticle enzyme immunoassay (MEIA). The urinary ratio of (6b-OHF þ 6b-OHE)/(F þ E) was calculated. Results Chromatography Figure 1A shows the representative chromatograms of standard compounds, Figure 1B shows a morning urine sample and Figure 1C shows the urine sample spiked with 6b-OHF (50 ng/ mL), 6b-OHE (50 ng/mL), E (50 ng/mL), F (40 ng/mL) and DM (50 ng/mL). The analytes, 6b-OHF, 6b-OHE, E, F and DM, were well separated under the previously described chromatographic conditions at retention times of 5.6, 6.7, 18.4, 19, 6 and 24.9 min, respectively. The peak shapes were acceptable and completely resolved one from another. No interfering peaks were observed in the urine samples. Linearity and sensitivity of the assay The peak area ratios of 6b-OHF, 6b-OHE, E and F to DM in human urine was linear with respect to the concentrations of the four analytes. A good correlation was found between the observed peak area ratios (y) and the theoretical concentration (x). Unweighted least-squares regression analysis provided typical regression lines: y ¼ 0.008184x þ 0.016877 (r 2 ¼ 0.99560) for 6b-OHF; y ¼ 0.021729x – 0.004091 (r 2 ¼ 0.99667) for 6b-OHE; y ¼ 0.012444x þ 0.038403 (r 2 ¼ 0.99767) for E; y ¼ 0.011752x þ 0.101080 (r 2 ¼ 0.99250) for F. When a signal-to-noise (S/N) ratio of 10.0 or greater was used as a criterion for a significant response, the lower limit of quantification (LLOQ) was established at 5 ng/mL for the four compounds. Extraction efficiency The extraction recovery values were determined at low, medium and high quality control samples of F, E, 6b-OHF and 6b-OHE. Extraction recovery values were also determined at a fixed concentration of DM (200 ng/mL). Results are shown in Table I.
Precision and accuracy of the assay The results shown in Tables II and III indicate that the assay method is reproducible for replicate analyses of E, F, 6b-OHF and 6b-OHE in human urine within the same day and on different days. The accuracy values for between-batch and within-batch studies at low, medium and high concentrations of E, F, 6b-OHF and 6b-OHE in urine were within acceptable limits (Tables II and III).
Figure 1. LC chromatograms of standard compounds (6b-OHF ¼ 400 ng/mL, 6b-OHE ¼ 400 ng/mL, E ¼ 400 ng/mL, F ¼ 320 ng/mL, DM ¼ 100 ng/mL) (A); morning urine sample (6b-OHF ¼ 113.64 ng/mL, 6b-OHE ¼ 22.92 ng/mL, E ¼ 105.56 ng/mL, F ¼ 70.00 ng/mL) (B); urine sample þ standard compounds þ DM (6b-OHF ¼ 130.59 ng/mL, 6b-OHE ¼ 50.27 ng/mL, E ¼ 122.22 ng/mL, F ¼ 50.84 ng/mL) (C).
Application of the method This method was applied to the urine samples collected from 30 renal transplant recipients; (6b-OHF þ 6b-OHE)/(F þ E), 6bOHF/F and 6b-OHE/E were 1.17 + 0.69, 2.76 + 2.36 and 0.51 + 0.84, respectively. The tacrolimus blood concentration/dose ratio (C/D) was 2.28 + 1.14 ng/mL. Correlations between the three ratios and C/D were examined by Spearman’s rank correlation analysis. All statistic analysis was performed by using SPSS 17.0 statistics software and a p value less than 0.05 was considered significant. A significant correlation was found between (6b-OHF þ 6b-OHE)/(F þ E) and C/D (r ¼ –0.824, p , 0.05).
Weaker correlations were also observed between 6b-OHF/F and C/D (r ¼ –0.511, p , 0.05), and 6b-OHE/E and C/D (r ¼ –0.098, p . 0.05). The values of C/D decreased while urinary (6b-OHF þ 6b-OHE)/(F þ E) increased.
Discussion Acetonitrile is effective to separate 6b-OHF and 6b-OHE, but for F and E, the degree of separation is poor. To obtain good separation and resolution, a solvent system composed of acetonitrile – water–methanol was selected as the mobile phase.
Simultaneous Determination of Cortisol, Cortisone, 6b-Hydroxycortisol and 6b-Hydroxycortisone by HPLC 453
Measured concentration (ng/mL)
Accuracy (%)
RSD (%)
9.85 + 0.32 49.08 + 3.29 200.04 + 2.56
98.53 98.16 100.02
3.28 6.70 1.28
11.54 + 1.09 51.58 + 4.39 204.24 + 2.63
115.47 103.11 102.12
9.45 8.52 1.29
losses of both metabolites, NaOH and acetic acid saturated with sodium sulfate, respectively, were applied to the organic extract. Additionally, to remove insoluble substances and minimize impurity peaks, urine samples were centrifuged at 12,000 rpm for 10 min before pretreatment. Generally speaking, the extraction recovery should be stable at high, medium and low concentrations, but the experiment presented here simultaneously determines four analytes, the properties of four analytes were considered synthetically when choosing an extracting agent. The low extraction recovery of F at a low concentration was probably caused by the limited solubility in the extracting agent, so the extracted quantity is limited. Subsequent washing of the organic extracts with alkaline and acidic solutions can also cause the loss of F; additionally, other urinary endogenous substances may influence the extraction, especially at a low concentration. Although the extraction recovery has certain fluctuations, it meets the standard for the analysis of biological samples. Based on a search of the literature, a few studies were reported on the search for tacrolimus metabolism. Kishino et al. (13) found a significant inverse correlation between 6b-OHF/F and C/D of tacrolimus in two liver transplant recipients (r ¼ 0.658, p , 0.05); however, the limitation of this study was the small number of patients. The results obtained in this study presented a more significant inverse relationship between (6b-OHF þ6b-OHE)/(F þ E) and C/D of tacrolimus (r ¼ –0.824, p , 0.05) in Chinese renal transplant recipients, greater than those between 6b-OHF/F and C/D and 6b-OHE/E and C/D, indicating that urinary (6b-OHF þ 6b-OHE)/(F þ E) may be a more appropriate representation for predicting the disposition of tacrolimus. Although immunochemical techniques are highly sensitive, the four endogenous compounds cannot be separated by this method due to the cross-reactivity of the antibodies, because the four structures are highly similar. Gas chromatography –mass spectrometry (MS) and LC– MS can provide acceptable specificity and sensitivity, but the use of these techniques is generally limited for routine analysis, primarily due to the availability of instruments. This paper described a reliable reversed-phase HPLC assay to simultaneously quantify E, F, 6b-OHF and 6b-OHE in human urine. Distinct advantages include good resolution between the parent drug and its metabolite, good resolution of the analytes from endogenous compounds, accurate assay of large numbers of samples and the requirement of only common instruments.
11.55 + 0.34 50.74 + 3.13 199.39 + 3.32
115.50 101.50 99.70
2.95 6.19 1.66
Acknowledgments
10.61 + 0.64 39.68 + 2.74 161.45 + 15.05
106.09 99.20 100.91
6.02 6.91 9.32
The research was funded by National Natural Science Foundation of China (81072700/H3110) and Doctoral Innovation Projects of Hunan Province (CX2011B055).
Table I Extraction Efficiency of 6b-OHF, 6b-OHE, F, E and DM for Urine Samples* Compound
Concentration (ng/mL)
Recovery (%)
RSD† (%)
6b-OHF
10 50 200 10 50 200 10 40 160 10 50 200 200
62.7 + 0.27 75.0 + 0.83 73.5 + 8.46 69.8 + 0.46 77.8 + 1.44 69.1 + 13.1 65.9 + 0.67 83.9 + 2.16 91.2 + 3.02 63.2 + 0.50 86.5 + 1.81 75.8 + 12.5 91.8 + 5.7
4.4 2.1 5.3 6.7 3.9 9.4 10.8 6.4 2.0 7.7 4.3 8.5 7.1
6b-OHE
F
E
DM
*Note: Data represent mean + standard deviation (SD); n ¼ 5. † Relative standard deviation.
Table II Intra-Day Precision and Accuracy of 6b-OHF, 6b-OHE, E and F in Urine Samples* Added concentration (ng/mL) 6b-OHF 10 50 200 6b-OHE 10 50 200 E 10 50 200 F 10 40 160
Measured concentration (ng/mL)
Accuracy (%)
RSD (%)
10.07 + 0.80 49.08 + 1.95 200.06 + 4.26
100.67 98.16 100.03
7.95 3.96 2.13
10.34 + 0.82 51.58 + 1.50 205.12 + 3.37
103.40 103.11 102.56
7.94 2.92 1.64
11.75 + 0.60 50.75 + 1.83 201.09 + 2.08
117.46 101.50 96.17
5.10 3.61 1.03
9.91 + 0.38 39.67 + 1.25 153.87 + 1.36
99.05 99.20 96.17
3.91 3.14 0.88
*Note: Data represent mean + SD; n ¼ 5.
Table III Inter-Day Precision and Accuracy of 6b-OHF, 6b-OHE, E and F in Urine Samples* Added concentration (ng/mL) 6b-OHF 10 50 200 6b-OHE 10 50 200 E 10 50 200 F 10 40 160
*Note: Data represent mean + SD; n ¼ 5.
Through repeated analysis, it was concluded that all compounds were completely separated when the gradient elution of methanol was 20 –34%. The water solubility of 6b-OHF and 6b-OHE is better than F, E and DM, so in the process of removing impurities, to cut the 454 Zheng et al.
References 1. Hane, K., Fujioka, M., Namiki, Y., Kitagawa, T., Kihara, N., Shimatani, K., et al.; Physico-chemical properties of FK-506; Iyakuhin Kenkyu, (1992); 23: 33–43. 2. Zheng, H., Webber, S., Zeevi, A., Schuetz, E., Zhang, J., Bowman, P., et al.; Tacrolimus dosing in pediatric heart transplant patients is
3.
4.
5.
6.
7.
8.
related to CYP3A5 and MDR1 gene polymorphisms; American Journal of Transplantation, (2003); 3: 477– 483. Iwasaki, K; Metabolism of tacrolimus (FK506) and recent topics in clinical pharmacokinetics; Drug Metabolism and Pharmacokinetics, (2007); 22: 328–335. Kuypers, D.R.; Immunosuppressive drug monitoring—What to use in clinical practice today to improve renal graft outcome; Transplant International, (2005); 18: 140– 150. Abel, S.M., Maggs, J.L., Back, D.J., Park, B.K.; Cortisol metabolism by human liver in vitro—I. Metabolite identification and inter-individual variability; Journal of Steroid Biochemistry and Molecular Biology, (1992); 43: 713–719. Luo, X., Zhu, L.J., Wu, W., Sheng, X.X., Cai, N.F., Liu, S.K., et al.; Simultaneous Determination of 6b-hydroxycortisol and 6b-hydroxycortisone in human urine by LC with UV absorbance detection; Chromatographia, (2009); 70: 1215–1219. Jin, M.J., Shimada, T., Yokogawa, K.; Contributions of intestinal P-glycoprotein and CYP3A to oral bioavailability of cyclosporin A in mice treated with or without dexamethasone; International Journal of Pharmaceutics, (2006); 309: 81– 86. Perogamvros, L., Owen, L.J., Newell-Price, J., Ray, D.W., Trainer, P.J., Keevil, B.G.; Simultaneous measurement of cortisol and cortisone in human saliva using liquid chromatography–tandem mass spectrometry: Application in basal and stimulated conditions; Journal of Chromatography B, (2009); 877: 3771–3775.
9. Santos-Montes, A., Gonzalo-Lumbreras, R., Izquierdo-Hornillos, R.; Simultaneous determination of cortisol and cortisone in urine by reversed-phase high-performance liquid chromatography clinical and doping control applications; Journal of Chromatography B, (1995); 673: 27 –33. 10. Goto, J., Shamsa, F., Goto, N., Nambara, T.; The simultaneous determination of serum cortisol and cortisone by highperformance liquid chromatography with fluorimetric detection; Journal of Pharmaceutical and Biomedical Analysis, (1983); 1: 83– 88. 11. Hu, Z.Y., Gong, Q., Hu, X.L., Wang, L.Q., Cao, Y.J., Cao, W.; Simultaneous determination of 6b-hydroxycortisol and cortisol in human, urine and plasma by liquid chromatography with ultraviolet absorbance detection for phenotyping the CYP3A activity; Journal of Chromatography B, (2005); 826: 238–243. 12. Lutz, U., Bittner, N., Ufer, M., Lutz, W.K.; Quantification of cortisol and 6 beta-hydroxycortisol in human urine by LC-MS/MS, and gender-specific evaluation of the metabolic ratio as biomarker of CYP3A activity; Journal of Chromatography B, (2010); 878: 97– 10. 13. Kishino, S., Ogawa, M., Akekuma, Y., Sugawara, M., Shimamura, T., Furukawa, H.; The variability of liver graft function and urinary 6b-hydroxycortisol to cortisol ratio during liver regeneration in liver transplant recipients; Clinical Transplantation, (2004); 18: 124– 129.
Simultaneous Determination of Cortisol, Cortisone, 6b-Hydroxycortisol and 6b-Hydroxycortisone by HPLC 455
