- Email: [email protected]
Determination of the antifungal agent posaconazole in human serum by HPLC with parallel column-switching technique
Journal of Chromatography B, 877 (2009) 2493–2498
Contents lists available at ScienceDirect
Journal of Chromatography B journal homepage: www.elsevier.com/locate/chromb
Determination of the antifungal agent posaconazole in human serum by HPLC with parallel column-switching technique Werner Neubauer a,b , Armin König c , Richard Bolek c , Rainer Trittler b , Monika Engelhardt a , Manfred Jung d , Klaus Kümmerer c,∗ a
Department of Hematology & Oncology, University Medical Center Freiburg, Freiburg, Germany Hospital Pharmacy, University Medical Center Freiburg, Freiburg, Germany Department of Environmental Health Sciences, University Medical Center Freiburg, Freiburg, Germany d Institute of Pharmaceutical Sciences, University of Freiburg, Freiburg, Germany b c
a r t i c l e
i n f o
Article history: Received 13 February 2009 Accepted 12 June 2009 Available online 21 June 2009 Keywords: Posaconazole Bioavailability Serum HPLC Column-switching On-line sample preparation
a b s t r a c t Posaconazole is a new broad-spectrum antifungal agent that is currently only available as an oral suspension and shows high intra- and inter-individual differences in oral bioavailibility. Pre-existing methods for the determination of the substance involve the use of internal standards or require a quite complicated and time-consuming sample pre-treatment. Our HPLC method is fast and fully-automated and there is no need for any manual sample pre-treatment. On-line transfer of posaconazole from the extraction column was followed by chromatographic separation on a C18 column and fluorescence detection (ex : 261 nm, em : 357 nm). Retention time of posaconazole was about 11.7 min, the lower limit of quantification was found to be 0.1 mg/l. A linear calibration curve was obtained over the concentration range of 0.1–5 mg/l using a 50 l sample (r2 = 0.999). The relative standard deviations of intra-day variations ranged from 2.3% to 9.4%, intra-day accuracy from 88.8% to 114.8%. © 2009 Elsevier B.V. All rights reserved.
1. Introduction Invasive fungal infections (IFIs) are associated with high morbidity and mortality rates among patients with hematological diseases and in those undergoing hematopoietic stem cell transplantation [1–3]. Their incidence has increased significantly during the last 20 years and the current treatment options provide limited benefit [1,4–6]. Posaconazole (4-4-[4-(4-{(3R,5R)-5-(2,4-difluorophenyl)-5(1H-1,2,4-triazol-1-ylmethyl)tetrahydro-3-furanyl}methoxyphenyl)piperazino]phenyl-1-[(1S,2S)-1-ethyl-2-hydroxypropyl]4,5-dihydro-1H-1,2,4-triazol-5-one) (chemical structure in Fig. 1) is a new antifungal triazole used for the prophylaxis and therapy of invasive fungal infections [7,8]. It is a derivative of itraconazole with a high in vitro and in vivo activity against a number of molds and yeasts, including Candida and Aspergillus species, but also less common fungi, such as the Zygomycetes, Cryptococcus or fusarium species [9–14]. As all triazole antimycotics, the substance works by inhibiting the cytochrome P450 dependent lanosterol
∗ Corresponding author at: Department of Environmental Health Sciences, University Medical Center Freiburg, Breisacher Str. 115B, D-79106 Freiburg i. Br., Germany. E-mail address: [email protected] (K. Kümmerer). 1570-0232/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.jchromb.2009.06.022
14-␣-demethylase, which catalyses an essential step in ergosterol biosynthesis. So far, posaconazole is not available as an intravenous formulation and has to be given as an oral suspension. Comparable to itraconazole, its bioavailability can be augmented when given with food (area under the curve [AUC] about four times greater when administered with high-fat meals) [15]. A daily administration of 800 mg, given in two divided doses, resulted in a maximum exposure, but there are high intra- and inter-individual differences [16,17]. Oral exposure, for instance, is especially low in patients after bone marrow or peripheral blood stem cell transplantation. The American product information reports on a positive association between average posaconazole levels in the blood and the efficacy of antifungal prophylaxis, and a possible association between an increased risk of treatment failure with lower posaconazole concentrations is described [7]. An open-label multicenter study recently reported that higher plasma concentrations were associated with improved response rates in the treatment of invasive aspergillosis [18]. For these reasons, monitoring serum posaconazole levels is recommended at least for patients with suspected poor oral absorption, patients with progressive disease under posaconazole therapy, patients on concomitant medications with significant drug interactions [19] and patients at risk for lower serum posaconazole levels (e.g. patients after bone marrow or peripheral blood stem cell transplantation).
2494
W. Neubauer et al. / J. Chromatogr. B 877 (2009) 2493–2498
Fig. 1. Chemical structure of posaconazole.
Previously published HPLC-based methods that are highly effective for the determination of serum posaconazole levels involve the use of internal standards and require a quite complicated and timeconsuming sample pre-treatment that increases the possibility of interferences. A method described by Chhun et al. [26] involves the addition of a mixture of hexane and methylene chloride that has to be evaporated to dryness afterwards which appears as a very complex and time-consuming procedure. Diazepam is used as internal standard in this method. The methods by Müller et al. and Kim et al. [17,27] involve the precipitation of serum proteins with acetonitrile or methanol, the first approach utilizing itraconazole as internal standard. Besides the possibility of interferences associated with sample pre-treatment, the use of internal standards is also related to analytical problems: not only the internal standard has to be absent in the patient sample, but also any other substance inducing a peak at the same retention time. Since cancer patients, and especially those receiving allogeneic or autologous stem cell transplants, receive a wide variety of concomitantly applied drugs (e.g. antibiotics, immunosuppressive agents), there is always a possibility of interferences that should not be underestimated. In this article, we present a fast and simple method for the determination of posaconazole levels in human serum based on high-performance liquid chromatography (HPLC) and fluorescence detection. Due to the use of parallel column-switching technique and an on-line extraction system leading to a separation of the serum proteins from the analyte by use of an extraction column, there is no need for any complex sample pre-treatment or the use of internal standards and the serum can be directly injected into the HPLC system. Besides the described methods, another analytical approach using mass spectrometry (MS) as the mode of detection was published by Shen et al. [28]. This method is both highly reliable and sensitive and showed remarkable results in the determination of posaconazole in human plasma. Yet, since not every hospital or laboratory has access to mass spectrometry and the associated costs exceed those of other procedures, methods based on fluorescence or UV detection are likely to be a more suitable alternative for the standard routine. 2. Experimental 2.1. Reagents Acetonitrile (LiChrosolv® , gradient grade) and formic acid (analytical grade) were purchased from Merck (Darmstadt, Germany). HPLC-grade water was generated using a Milli-Q water-purification system from Millipore (Molsheim, France). Pooled blank serum samples were obtained from Freiburg University Medical Center. Posaconazole was kindly obtained from the Chemical Research Division, Schering-Plough Research Institute (Kenilworth, NJ, USA).
2.2. Calibration standards The posaconazole stock solution had a concentration of 100 mg/l and was dissolved in an aqueous solution of formic acid (1.5%). Dissolution in water was not possible due to the poor solubility of the substance. The posaconazole dilutions (0.1, 0.25, 0.5, 1.0, 2.5 and 5.0 mg/l) and pooled blank serum samples were stored at −20 ◦ C until analysis. 2.3. Patient samples Blood samples were obtained from inpatients of the Hematology & Oncology Department who received posaconazole for therapeutic or prophylactic reasons. The blood was drawn before the intake of posaconazole as a morning trough value and was centrifuged (Universal 32 R, Hettich Zentrifugen, Tuttlingen, Germany) at 5000 g for 10 min to obtain the pure serum. The study design was approved by the ethics committee of the University of Freiburg. 2.4. Liquid chromatography HPLC analysis was performed by using an integrated Shimadzu HPLC-system (Shimadzu Deutschland GmbH, Duisburg, Germany) consisting of four pumps (Shimadzu LC-10AT, LC-10AD and LC10AS), two six-port switching valves (Shimadzu FCV-12AH) and a communications bus module (Shimadzu CBM-10A). A Uniflow Degasys DG 1310 degasser (VDS optilab, Berlin, Germany) was also integrated into the system. The configuration of the HPLC-system is shown schematically in Fig. 2A. The column oven (Shimadzu CTO-10AC) was set to a constant temperature of 40 ◦ C. The autosampler (Shimadzu SIL-10ADVP) and pump 1A were used to load 50 l of the serum samples containing posaconazole onto the extraction column (ADS) for the separation of the analyte from the serum proteins. The mobile phase of pump 1A was formic acid 0.1% in water (v/v) and was delivered to the ADS extraction column at a flow rate of 1 ml/min. Posaconazole was retained on the extraction column, while matrix compounds were flushed to waste with the eluent. After 5 min, the matrix had been fully washed out of the extraction column. Simultaneously, the analytical column (C18 ) was equilibrated by pumps 2A and B with formic acid (0.1%) in water (v/v) and acetonitrile in a ratio of 55:45 (v/v) at a flow rate of 1 ml/min. The software time-schedule then automatically switched the high-pressure valve 2 into transfer position (Fig. 2B), thereby coupling the extraction column with the analytical column. The analytical mobile phase was delivered from pump 2A and B in a ratio of 55:45 (v/v). Within 5 min, posaconazole was rapidly eluted from the extraction column by back flushing at a flow rate of 1 ml/min. The higher elution power of acetonitrile desorbed the analyte from the extraction column and transferred it to the analytical column.
W. Neubauer et al. / J. Chromatogr. B 877 (2009) 2493–2498
2495
Fig. 2. (A) Scheme of the HPLC-integrated sample preparation – sample injection step: the analytical column is separated from the extraction side. (B) Scheme of the HPLC-integrated sample preparation – transfer and chromatographic separation step: the extraction column is connected with the analytical column.
Chromatography itself was then performed on a 125/4 Nucleodur® 100-5 C18ec analytical column preceded by a guard column 8/4 100-5 C18ec (both from Macherey & Nagel, Düren, Germany). The chromatographic separation on the analytical column was accomplished by an isocratic mobile phase gradient of 55:45 (v/v) at a flow rate of 1 ml/min and detection was performed by usage of a fluorescence detector (RF-10A, excitation wavelength: 261 nm, emission wavelength: 357 nm). Simultaneously, the autosampler was rinsed with pure acetonitrile from pump 1B at a flow rate of 2 ml/min. 10 min after
having started the analysis, the high-pressure valve 2 switched back to the initial position. The extraction column was now rinsed with acetonitrile from pump 1B (flow rate: 2 ml/min) while chromatographic separation on the analytical column and detection with the fluorescence detector were performed concomitantly. The initial conditions were restored after 14 min, sum of sample preparation and analysis time was 17 min. After that, a new sample could be injected and extraction started, while chromatographic separation of the former sample was finished simultaneously.
2496
W. Neubauer et al. / J. Chromatogr. B 877 (2009) 2493–2498
Fig. 3. Chromatograms of blank human serum (A) and human serum spiked with 0.1 mg/l posaconazole (B). Peaks after 6 and 7 min due to column-switching.
Of note, valve 1 is not imperatively necessary since it is not switched during analysis. It is therefore also possible to directly connect the line of pumps 1A and 1B and the autosampler to port 6 of valve 2 and the line of pumps 2A and 2B to port 3 of valve 2. The method is described with two switching valves since this configuration is well-established in our institution and provides high flexibility for the analysis of various drugs, e.g. for the determination of the highly absorptive antifungal drug caspofungin for which two switching valves are required [25]. 3. Results and discussion 3.1. Selectivity No endogenous components extracted from blank serum eluted at the retention time of posaconazole (Fig. 3). At least five blank plasma samples were assayed in order to investigate possible interferences with the matrix. Furthermore, no interferences with other drugs could be detected at the retention time of posaconazole (11.7 min). For this purpose, we injected n = 22 plasma samples of eight patients not receiving posaconazole, but a wide range of drugs that are quite commonly applied in patients being treated for hematological or oncological diseases (i.e. vancomycin, phenytoin, cyclosporine, aciclovir, ganciclovir, voriconazole, clarithromycin, teicoplanin, meropenem and metoclopramide). 3.2. Linearity Linearity of the calibration curve in serum was tested over the range of 0.1–5 mg/l (6 data points, 10 measurements per data point).
The described method was linear over the tested concentrations with a correlation coefficient of r2 = 0.999. 3.3. Lower limit of quantification and limit of detection The lower limit of quantification (LLOQ) was about 0.1 mg/l and was determined when the accuracy was within 80–120% and the relative standard deviation was within 20%. For this method, the LLOQ was found to be 0.1 mg/l with a relative standard deviation (RSD) of precision of 6.0% and accuracy of 83.7%. The limit of detection (LOD) was about 0.05 mg/l and was defined as the lowest quantity of analyte determined with an analyte signal at least three times the noise of the blank response. The sensitivity of this method is supposed to be sufficient for the clinical use, since reported therapeutic concentrations (therapeutic dosage: 800 mg posaconazole per day) range between 0.4 and 2 mg/l [16]. 3.4. Accuracy and precision Four concentrations (0.25, 0.5, 1.0 and 2.5 mg/l) in ten replicates were utilized to validate the accuracy and precision of the developed method. Inter-day validation was assessed by measuring the indicated concentrations in five replicates on two consecutive days, respectively. Intra-day accuracy ranged between 88.8% and 114.8%, inter-day accuracy between 89.2% and 109.9%. The highest RSD was 9.4% for intra-day precision and 9.6% for inter-day precision (Table 1). Most validation data of our method are comparable with those of the method reported by Chhun et al. [26] that includes the use of diazepam as internal standard and a complex and time-consuming
Table 1 Intra- and inter-day precision and accuracy of HPLC assay for the determination of posaconazole in human serum. Spiked concentration (mg/l) 0.25
0.50
1.00
2.50
Batch 1 (n = 5) Observed intra-day mean (mg/l) Intra-day precision (%) Intra-day accuracy (%)
0.222 ± 0.011 4.8 88.8
0.468 ± 0.017 3.7 93.6
1.049 ± 0.024 2.3 104.9
2.392 ± 0.094 3.9 95.7
Batch 2 (n = 5) Observed intra-day mean (mg/l) Intra-day precision (%) Intra-day accuracy (%)
0.224 ± 0.011 4.9 89.6
0.497 ± 0.047 9.4 99.4
1.148 ± 0.028 2.4 114.8
2.455 ± 0.089 3.6 98.2
Inter-day (n = 10) Observed inter-day mean (mg/l) Inter-day precision (%) Inter-day accuracy (%)
0.223 ± 0.009 4.2 89.2
0.482 ± 0.046 9.6 96.5
1.099 ± 0.019 1.8 109.9
2.424 ± 0.046 1.9 96.9
W. Neubauer et al. / J. Chromatogr. B 877 (2009) 2493–2498
2497
sample pre-treatment. For instance, the lower limit of quantification, the intra- and inter-day precision and the mean recoveries are in the same range for both methods. Concerning accuracy and lower limit of quantification, data are slightly better with the method developed by Chhun et al. However, our validation data are sufficient to obtain reliable results and our procedure is less complex and therefore clearly more practical in routine clinical use. 3.5. Recovery The calibration curve in serum was virtually identical to a calibration curve obtained with posaconazole dissolved in pure acetonitrile with an average recovery of more than 97% (data not shown). 3.6. Method development: optimization of essential steps Concerning the characteristics of the extraction column, it was necessary to use column packing materials with an adequate capacity to retain the drug and therefore to allow a preferably quantitative separation of the analyte and matrix components. For this reason, restricted access column packing materials (LiChrospher® RP-8 ADS, Merck KG, Darmstadt, Germany) were utilized which allow the direct and repetitive injection of untreated serum samples [20,21]. Due to the preceding separation of matrix components by means of an extraction column, the damage to the analytical column was not very serious. So far, the column was utilized for the analysis of more than 200 unprocessed serum samples with the same performance and a quite constant serum peak. Restricted access materials consist of spherical particles with a diameter of 25 m and two chemically different surfaces: the chemically inert outer surface includes hydrophilic, electroneutral diol groups and protects the column from any unwanted contamination caused by interaction with the protein matrix. The inner surface of the porous particles with restricted access is covered with a hydrophobic dispersion phase (C4 , C8 , C18 alkyl chains) that has adsorptive properties and is accessible to low molecular analytes only. Such column packing materials have been used successfully for the analysis of antibiotics and other antimycotics in serum [22–24]. To optimize the washing step of the extraction column, several solvents were tested. These comprised variably concentrated solutions of formic acid in water (0.05–0.5%, v/v) as well as mixtures of formic acid in water (0.1%, v/v) and acetonitrile with proportions of acetonitrile of 0–50%. Besides, different flow rates of the solvent (0.5–3 ml/min) and different durations of the sample injection step (Fig. 2A) and the transfer and chromatographic separation step (Fig. 2B) were applied in order to achieve optimal conditions for the complete elution of the serum matrix in the shortest possible time. Higher flow-rates and the addition of acetonitrile to the solvent both lead to an incomplete retention of posaconazole on the extraction column and were therefore not feasible whereas modifying the concentrations of formic acid in water resulted in an incomplete elution of the serum matrix. Instead of pre-mixed mobile phases, we used two different pumps (2A and 2B) for the delivery of the analytical mobile phase. This might appear complicated at the first view, but helped to simplify changes in the method, especially in the composition of the two components. 3.7. Analysis of patient samples Fig. 4 displays the course of posaconazole concentrations in a patient under chemotherapy for follicular lymphoma who had already received an autologous haematopoietic stem cell trans-
Fig. 4. Time course of serum concentrations of posaconazole in a patient at the end of treatment (for details see text).
Fig. 5. Posaconazole concentrations in relation to the applied doses (samples originating from five patients, n = 16).
plant. The patient came from another hospital with the unusual dosage of 400 mg posaconazole three times daily. We examined the serum concentrations over 5 days after discontinuation of the drug. The average half-life of posaconazole can be derived from the measured data to be approximately 35 h (Fig. 4). Fig. 5 shows posaconazole concentrations in relation to the applied doses (400, 600, 800 and 1200 mg daily) in five cancer patients. All blood samples (n = 16) were drawn before the intake of posaconazole as a morning trough value. The measured drug levels varied between 0.117 and 2.252 mg/l. It is obvious that there is a wide variability within the determined serum concentrations of the 800 mg/d dosage group. 4. Conclusion The presented HPLC method allows the fast and automated determination of the antifungal drug posaconazole in human serum without the need for deproteinization and thereby loss of analyte or any other form of sample pre-treatment. The first step of the analysis is an automated sample clean-up by use of an extraction column packed with a restricted access material which leads to the separation of the analyte from the serum proteins and allows the direct and repetitive injection of untreated serum samples. The second step consists of the chromatographic separation performed on an RP18 analytical column followed by identification with a fluorescence detector. Several pre-existing methods for the determination of posaconazole in human serum report slightly lower LOD or LLOQ [17,27], but the sensitivity and selectivity of our method are absolutely sufficient for routine clinical use. Furthermore, as not every hospital or laboratory has access to mass spectrometry, methods
2498
W. Neubauer et al. / J. Chromatogr. B 877 (2009) 2493–2498
based on fluorescence detection are likely to be a more suitable alternative for the standard routine. Acknowledgements We thank the Chemical Research Division, Schering-Plough Research Institute (Kenilworth, NJ, USA) for supplying the pure substance of posaconazole and the central medical chemistry laboratory of Freiburg University Medical Center for the pooled blank serum. We are also obliged to Christoph Trautwein, Armin Schuster and Andreas Längin for their help and technical assistance and to Egid Strehl for his generous support. References [1] K.A. Marr, R.A. Carter, F. Crippa, A. Wald, L. Corey, Clin. Infect. Dis 34 (2002) 909. [2] Z. Bhatti, A. Shaukat, N.G. Almyroudis, B.H. Segal, Mycopathologia 162 (2006) 1. [3] E.J. Bow, Br. J. Haematol. 140 (2008) 133. [4] M.A. Pfaller, D.J. Diekema, J. Clin. Microbiol. 42 (2004) 4419. [5] P.G. Pappas, J.H. Rex, J. Lee, R.J. Hamill, R.A. Larsen, W. Powderly, C.A. Kauffman, N. Hyslop, J.E. Mangino, S. Chapman, H.W. Horowitz, J.E. Edwards, W.E. Dismukes, Clin. Infect. Dis. 37 (2003) 634. [6] R. Herbrecht, D.W. Denning, T.F. Patterson, J.E. Bennett, R.E. Greene, J.W. Oestmann, W.V. Kern, K.A. Marr, P. Ribaud, O. Lortholary, R. Sylvester, R.H. Rubin, J.R. Wingard, P. Stark, C. Durand, D. Caillot, E. Thiel, P.H. Chandrasekar, M.R. Hodges, H.T. Schlamm, P.F. Troke, B. de Pauw, N. Engl. J. Med. 347 (2002) 408. [7] Noxafil (posaconazole), Prescribing Information, Schering Corporation, Kenilworth, NJ, October 2006 (USA). [8] Noxafil (posaconazole), European Prescribing Information, Schering-Plough Europe, Brussels, Belgium, March 2007 (European Union). [9] R. Herbrecht, Int. J. Clin. Pract. 58 (2004) 612.
[10] G.M. Keating, Drugs 65 (2005) 1553. [11] J.A. Vazquez, D.J. Skiest, L. Nieto, R. Northland, I. Sanne, J. Gogate, W. Greaves, R. Isaacs, Clin. Infect. Dis. 42 (2006) 1179. [12] R.N. Greenberg, K. Mullane, J.A. van Burik, I. Raad, M.J. Abzug, G. Anstead, R. Herbrecht, A. Langston, K.A. Marr, G. Schiller, M. Schuster, J.R. Wingard, C.E. Gonzalez, S.G. Revankar, G. Corcoran, R.J. Kryscio, R. Hare, Antimicrob. Agents Chemother. 50 (2006) 126. [13] F. Barchiesi, D. Arzeni, A.W. Fothergill, L.F. Di Francesco, F. Caselli, M.G. Rinaldi, G. Scalise, Antimicrob. Agents Chemother. 44 (2000) 226. [14] K.L. Oakley, C.B. Moore, D.W. Denning, Antimicrob. Agents Chemother. 41 (1997) 1124. [15] R. Courtney, S. Pai, M. Laughlin, J. Lim, V. Batra, Antimicrob. Agents Chemother. 47 (2003) 2788. [16] A.J. Ullmann, O.A. Cornely, A. Burchardt, R. Hachem, D.P. Kontoyiannis, K. Töpelt, R. Courtney, D. Wexler, G. Krishna, M. Martinho, G. Corcoran, I. Raad, Antimicrob. Agents Chemother. 50 (2006) 658. [17] C. Müller, M. Arndt, C. Queckenberg, O.A. Cornely, M. Theisohn, Mycoses 49 (2006) 17. [18] T.J. Walsh, I. Raad, T.F. Patterson, P. Chandrasekar, G.R. Donowitz, R. Graybill, R.E. Greene, R. Hachem, S. Hadley, R. Herbrecht, A. Langston, A. Louie, P. Ribaud, B.H. Segal, D.A. Stevens, J.A. van Burik, C.S. White, G. Corcoran, J. Gogate, G. Krishna, L. Pedicone, C. Hardalo, J.R. Perfect, Clin. Infect. Dis. 44 (2007) 2. [19] M.L. Goodwin, R.H. Drew, J. Antimicrob. Chemother. 61 (2008) 17. [20] K. Boos, C. Grimm, Trends Analyt. Chem. 18 (1999) 175. [21] R. Trittler, M. Ehrlich, T.J. Galla, R.E. Horch, K. Kümmerer, J. Chromatogr. B 775 (2002) 127. [22] M. Ehrlich, R. Trittler, F.D. Daschner, K. Kümmerer, J. Chromatogr. B 755 (2001) 373. [23] H. Egle, R. Trittler, A. König, K. Kümmerer, J. Chromatogr. B 814 (2005) 361. [24] H. Egle, R. Trittler, K. Kümmerer, Ther. Drug Monit. 26 (2004) 425. [25] H. Egle, R. Trittler, K. Kümmerer, Rapid Commun Mass Spectrom. 18 (2004) 2871. [26] S. Chhun, E. Rey, A. Tran, O. Lortholary, G. Pons, V. Jullien, J. Chromatogr. B 852 (2007) 223. [27] H. Kim, P. Likhari, P. Kumari, C.C. Lin, A.A. Nomeir, J. Chromatogr. B 738 (2000) 93. [28] J.X. Shen, G. Krishna, R.N. Hayes, J. Pharm. Biomed. Anal. 43 (2007) 228.
Contents lists available at ScienceDirect
Journal of Chromatography B journal homepage: www.elsevier.com/locate/chromb
Determination of the antifungal agent posaconazole in human serum by HPLC with parallel column-switching technique Werner Neubauer a,b , Armin König c , Richard Bolek c , Rainer Trittler b , Monika Engelhardt a , Manfred Jung d , Klaus Kümmerer c,∗ a
Department of Hematology & Oncology, University Medical Center Freiburg, Freiburg, Germany Hospital Pharmacy, University Medical Center Freiburg, Freiburg, Germany Department of Environmental Health Sciences, University Medical Center Freiburg, Freiburg, Germany d Institute of Pharmaceutical Sciences, University of Freiburg, Freiburg, Germany b c
a r t i c l e
i n f o
Article history: Received 13 February 2009 Accepted 12 June 2009 Available online 21 June 2009 Keywords: Posaconazole Bioavailability Serum HPLC Column-switching On-line sample preparation
a b s t r a c t Posaconazole is a new broad-spectrum antifungal agent that is currently only available as an oral suspension and shows high intra- and inter-individual differences in oral bioavailibility. Pre-existing methods for the determination of the substance involve the use of internal standards or require a quite complicated and time-consuming sample pre-treatment. Our HPLC method is fast and fully-automated and there is no need for any manual sample pre-treatment. On-line transfer of posaconazole from the extraction column was followed by chromatographic separation on a C18 column and fluorescence detection (ex : 261 nm, em : 357 nm). Retention time of posaconazole was about 11.7 min, the lower limit of quantification was found to be 0.1 mg/l. A linear calibration curve was obtained over the concentration range of 0.1–5 mg/l using a 50 l sample (r2 = 0.999). The relative standard deviations of intra-day variations ranged from 2.3% to 9.4%, intra-day accuracy from 88.8% to 114.8%. © 2009 Elsevier B.V. All rights reserved.
1. Introduction Invasive fungal infections (IFIs) are associated with high morbidity and mortality rates among patients with hematological diseases and in those undergoing hematopoietic stem cell transplantation [1–3]. Their incidence has increased significantly during the last 20 years and the current treatment options provide limited benefit [1,4–6]. Posaconazole (4-4-[4-(4-{(3R,5R)-5-(2,4-difluorophenyl)-5(1H-1,2,4-triazol-1-ylmethyl)tetrahydro-3-furanyl}methoxyphenyl)piperazino]phenyl-1-[(1S,2S)-1-ethyl-2-hydroxypropyl]4,5-dihydro-1H-1,2,4-triazol-5-one) (chemical structure in Fig. 1) is a new antifungal triazole used for the prophylaxis and therapy of invasive fungal infections [7,8]. It is a derivative of itraconazole with a high in vitro and in vivo activity against a number of molds and yeasts, including Candida and Aspergillus species, but also less common fungi, such as the Zygomycetes, Cryptococcus or fusarium species [9–14]. As all triazole antimycotics, the substance works by inhibiting the cytochrome P450 dependent lanosterol
∗ Corresponding author at: Department of Environmental Health Sciences, University Medical Center Freiburg, Breisacher Str. 115B, D-79106 Freiburg i. Br., Germany. E-mail address: [email protected] (K. Kümmerer). 1570-0232/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.jchromb.2009.06.022
14-␣-demethylase, which catalyses an essential step in ergosterol biosynthesis. So far, posaconazole is not available as an intravenous formulation and has to be given as an oral suspension. Comparable to itraconazole, its bioavailability can be augmented when given with food (area under the curve [AUC] about four times greater when administered with high-fat meals) [15]. A daily administration of 800 mg, given in two divided doses, resulted in a maximum exposure, but there are high intra- and inter-individual differences [16,17]. Oral exposure, for instance, is especially low in patients after bone marrow or peripheral blood stem cell transplantation. The American product information reports on a positive association between average posaconazole levels in the blood and the efficacy of antifungal prophylaxis, and a possible association between an increased risk of treatment failure with lower posaconazole concentrations is described [7]. An open-label multicenter study recently reported that higher plasma concentrations were associated with improved response rates in the treatment of invasive aspergillosis [18]. For these reasons, monitoring serum posaconazole levels is recommended at least for patients with suspected poor oral absorption, patients with progressive disease under posaconazole therapy, patients on concomitant medications with significant drug interactions [19] and patients at risk for lower serum posaconazole levels (e.g. patients after bone marrow or peripheral blood stem cell transplantation).
2494
W. Neubauer et al. / J. Chromatogr. B 877 (2009) 2493–2498
Fig. 1. Chemical structure of posaconazole.
Previously published HPLC-based methods that are highly effective for the determination of serum posaconazole levels involve the use of internal standards and require a quite complicated and timeconsuming sample pre-treatment that increases the possibility of interferences. A method described by Chhun et al. [26] involves the addition of a mixture of hexane and methylene chloride that has to be evaporated to dryness afterwards which appears as a very complex and time-consuming procedure. Diazepam is used as internal standard in this method. The methods by Müller et al. and Kim et al. [17,27] involve the precipitation of serum proteins with acetonitrile or methanol, the first approach utilizing itraconazole as internal standard. Besides the possibility of interferences associated with sample pre-treatment, the use of internal standards is also related to analytical problems: not only the internal standard has to be absent in the patient sample, but also any other substance inducing a peak at the same retention time. Since cancer patients, and especially those receiving allogeneic or autologous stem cell transplants, receive a wide variety of concomitantly applied drugs (e.g. antibiotics, immunosuppressive agents), there is always a possibility of interferences that should not be underestimated. In this article, we present a fast and simple method for the determination of posaconazole levels in human serum based on high-performance liquid chromatography (HPLC) and fluorescence detection. Due to the use of parallel column-switching technique and an on-line extraction system leading to a separation of the serum proteins from the analyte by use of an extraction column, there is no need for any complex sample pre-treatment or the use of internal standards and the serum can be directly injected into the HPLC system. Besides the described methods, another analytical approach using mass spectrometry (MS) as the mode of detection was published by Shen et al. [28]. This method is both highly reliable and sensitive and showed remarkable results in the determination of posaconazole in human plasma. Yet, since not every hospital or laboratory has access to mass spectrometry and the associated costs exceed those of other procedures, methods based on fluorescence or UV detection are likely to be a more suitable alternative for the standard routine. 2. Experimental 2.1. Reagents Acetonitrile (LiChrosolv® , gradient grade) and formic acid (analytical grade) were purchased from Merck (Darmstadt, Germany). HPLC-grade water was generated using a Milli-Q water-purification system from Millipore (Molsheim, France). Pooled blank serum samples were obtained from Freiburg University Medical Center. Posaconazole was kindly obtained from the Chemical Research Division, Schering-Plough Research Institute (Kenilworth, NJ, USA).
2.2. Calibration standards The posaconazole stock solution had a concentration of 100 mg/l and was dissolved in an aqueous solution of formic acid (1.5%). Dissolution in water was not possible due to the poor solubility of the substance. The posaconazole dilutions (0.1, 0.25, 0.5, 1.0, 2.5 and 5.0 mg/l) and pooled blank serum samples were stored at −20 ◦ C until analysis. 2.3. Patient samples Blood samples were obtained from inpatients of the Hematology & Oncology Department who received posaconazole for therapeutic or prophylactic reasons. The blood was drawn before the intake of posaconazole as a morning trough value and was centrifuged (Universal 32 R, Hettich Zentrifugen, Tuttlingen, Germany) at 5000 g for 10 min to obtain the pure serum. The study design was approved by the ethics committee of the University of Freiburg. 2.4. Liquid chromatography HPLC analysis was performed by using an integrated Shimadzu HPLC-system (Shimadzu Deutschland GmbH, Duisburg, Germany) consisting of four pumps (Shimadzu LC-10AT, LC-10AD and LC10AS), two six-port switching valves (Shimadzu FCV-12AH) and a communications bus module (Shimadzu CBM-10A). A Uniflow Degasys DG 1310 degasser (VDS optilab, Berlin, Germany) was also integrated into the system. The configuration of the HPLC-system is shown schematically in Fig. 2A. The column oven (Shimadzu CTO-10AC) was set to a constant temperature of 40 ◦ C. The autosampler (Shimadzu SIL-10ADVP) and pump 1A were used to load 50 l of the serum samples containing posaconazole onto the extraction column (ADS) for the separation of the analyte from the serum proteins. The mobile phase of pump 1A was formic acid 0.1% in water (v/v) and was delivered to the ADS extraction column at a flow rate of 1 ml/min. Posaconazole was retained on the extraction column, while matrix compounds were flushed to waste with the eluent. After 5 min, the matrix had been fully washed out of the extraction column. Simultaneously, the analytical column (C18 ) was equilibrated by pumps 2A and B with formic acid (0.1%) in water (v/v) and acetonitrile in a ratio of 55:45 (v/v) at a flow rate of 1 ml/min. The software time-schedule then automatically switched the high-pressure valve 2 into transfer position (Fig. 2B), thereby coupling the extraction column with the analytical column. The analytical mobile phase was delivered from pump 2A and B in a ratio of 55:45 (v/v). Within 5 min, posaconazole was rapidly eluted from the extraction column by back flushing at a flow rate of 1 ml/min. The higher elution power of acetonitrile desorbed the analyte from the extraction column and transferred it to the analytical column.
W. Neubauer et al. / J. Chromatogr. B 877 (2009) 2493–2498
2495
Fig. 2. (A) Scheme of the HPLC-integrated sample preparation – sample injection step: the analytical column is separated from the extraction side. (B) Scheme of the HPLC-integrated sample preparation – transfer and chromatographic separation step: the extraction column is connected with the analytical column.
Chromatography itself was then performed on a 125/4 Nucleodur® 100-5 C18ec analytical column preceded by a guard column 8/4 100-5 C18ec (both from Macherey & Nagel, Düren, Germany). The chromatographic separation on the analytical column was accomplished by an isocratic mobile phase gradient of 55:45 (v/v) at a flow rate of 1 ml/min and detection was performed by usage of a fluorescence detector (RF-10A, excitation wavelength: 261 nm, emission wavelength: 357 nm). Simultaneously, the autosampler was rinsed with pure acetonitrile from pump 1B at a flow rate of 2 ml/min. 10 min after
having started the analysis, the high-pressure valve 2 switched back to the initial position. The extraction column was now rinsed with acetonitrile from pump 1B (flow rate: 2 ml/min) while chromatographic separation on the analytical column and detection with the fluorescence detector were performed concomitantly. The initial conditions were restored after 14 min, sum of sample preparation and analysis time was 17 min. After that, a new sample could be injected and extraction started, while chromatographic separation of the former sample was finished simultaneously.
2496
W. Neubauer et al. / J. Chromatogr. B 877 (2009) 2493–2498
Fig. 3. Chromatograms of blank human serum (A) and human serum spiked with 0.1 mg/l posaconazole (B). Peaks after 6 and 7 min due to column-switching.
Of note, valve 1 is not imperatively necessary since it is not switched during analysis. It is therefore also possible to directly connect the line of pumps 1A and 1B and the autosampler to port 6 of valve 2 and the line of pumps 2A and 2B to port 3 of valve 2. The method is described with two switching valves since this configuration is well-established in our institution and provides high flexibility for the analysis of various drugs, e.g. for the determination of the highly absorptive antifungal drug caspofungin for which two switching valves are required [25]. 3. Results and discussion 3.1. Selectivity No endogenous components extracted from blank serum eluted at the retention time of posaconazole (Fig. 3). At least five blank plasma samples were assayed in order to investigate possible interferences with the matrix. Furthermore, no interferences with other drugs could be detected at the retention time of posaconazole (11.7 min). For this purpose, we injected n = 22 plasma samples of eight patients not receiving posaconazole, but a wide range of drugs that are quite commonly applied in patients being treated for hematological or oncological diseases (i.e. vancomycin, phenytoin, cyclosporine, aciclovir, ganciclovir, voriconazole, clarithromycin, teicoplanin, meropenem and metoclopramide). 3.2. Linearity Linearity of the calibration curve in serum was tested over the range of 0.1–5 mg/l (6 data points, 10 measurements per data point).
The described method was linear over the tested concentrations with a correlation coefficient of r2 = 0.999. 3.3. Lower limit of quantification and limit of detection The lower limit of quantification (LLOQ) was about 0.1 mg/l and was determined when the accuracy was within 80–120% and the relative standard deviation was within 20%. For this method, the LLOQ was found to be 0.1 mg/l with a relative standard deviation (RSD) of precision of 6.0% and accuracy of 83.7%. The limit of detection (LOD) was about 0.05 mg/l and was defined as the lowest quantity of analyte determined with an analyte signal at least three times the noise of the blank response. The sensitivity of this method is supposed to be sufficient for the clinical use, since reported therapeutic concentrations (therapeutic dosage: 800 mg posaconazole per day) range between 0.4 and 2 mg/l [16]. 3.4. Accuracy and precision Four concentrations (0.25, 0.5, 1.0 and 2.5 mg/l) in ten replicates were utilized to validate the accuracy and precision of the developed method. Inter-day validation was assessed by measuring the indicated concentrations in five replicates on two consecutive days, respectively. Intra-day accuracy ranged between 88.8% and 114.8%, inter-day accuracy between 89.2% and 109.9%. The highest RSD was 9.4% for intra-day precision and 9.6% for inter-day precision (Table 1). Most validation data of our method are comparable with those of the method reported by Chhun et al. [26] that includes the use of diazepam as internal standard and a complex and time-consuming
Table 1 Intra- and inter-day precision and accuracy of HPLC assay for the determination of posaconazole in human serum. Spiked concentration (mg/l) 0.25
0.50
1.00
2.50
Batch 1 (n = 5) Observed intra-day mean (mg/l) Intra-day precision (%) Intra-day accuracy (%)
0.222 ± 0.011 4.8 88.8
0.468 ± 0.017 3.7 93.6
1.049 ± 0.024 2.3 104.9
2.392 ± 0.094 3.9 95.7
Batch 2 (n = 5) Observed intra-day mean (mg/l) Intra-day precision (%) Intra-day accuracy (%)
0.224 ± 0.011 4.9 89.6
0.497 ± 0.047 9.4 99.4
1.148 ± 0.028 2.4 114.8
2.455 ± 0.089 3.6 98.2
Inter-day (n = 10) Observed inter-day mean (mg/l) Inter-day precision (%) Inter-day accuracy (%)
0.223 ± 0.009 4.2 89.2
0.482 ± 0.046 9.6 96.5
1.099 ± 0.019 1.8 109.9
2.424 ± 0.046 1.9 96.9
W. Neubauer et al. / J. Chromatogr. B 877 (2009) 2493–2498
2497
sample pre-treatment. For instance, the lower limit of quantification, the intra- and inter-day precision and the mean recoveries are in the same range for both methods. Concerning accuracy and lower limit of quantification, data are slightly better with the method developed by Chhun et al. However, our validation data are sufficient to obtain reliable results and our procedure is less complex and therefore clearly more practical in routine clinical use. 3.5. Recovery The calibration curve in serum was virtually identical to a calibration curve obtained with posaconazole dissolved in pure acetonitrile with an average recovery of more than 97% (data not shown). 3.6. Method development: optimization of essential steps Concerning the characteristics of the extraction column, it was necessary to use column packing materials with an adequate capacity to retain the drug and therefore to allow a preferably quantitative separation of the analyte and matrix components. For this reason, restricted access column packing materials (LiChrospher® RP-8 ADS, Merck KG, Darmstadt, Germany) were utilized which allow the direct and repetitive injection of untreated serum samples [20,21]. Due to the preceding separation of matrix components by means of an extraction column, the damage to the analytical column was not very serious. So far, the column was utilized for the analysis of more than 200 unprocessed serum samples with the same performance and a quite constant serum peak. Restricted access materials consist of spherical particles with a diameter of 25 m and two chemically different surfaces: the chemically inert outer surface includes hydrophilic, electroneutral diol groups and protects the column from any unwanted contamination caused by interaction with the protein matrix. The inner surface of the porous particles with restricted access is covered with a hydrophobic dispersion phase (C4 , C8 , C18 alkyl chains) that has adsorptive properties and is accessible to low molecular analytes only. Such column packing materials have been used successfully for the analysis of antibiotics and other antimycotics in serum [22–24]. To optimize the washing step of the extraction column, several solvents were tested. These comprised variably concentrated solutions of formic acid in water (0.05–0.5%, v/v) as well as mixtures of formic acid in water (0.1%, v/v) and acetonitrile with proportions of acetonitrile of 0–50%. Besides, different flow rates of the solvent (0.5–3 ml/min) and different durations of the sample injection step (Fig. 2A) and the transfer and chromatographic separation step (Fig. 2B) were applied in order to achieve optimal conditions for the complete elution of the serum matrix in the shortest possible time. Higher flow-rates and the addition of acetonitrile to the solvent both lead to an incomplete retention of posaconazole on the extraction column and were therefore not feasible whereas modifying the concentrations of formic acid in water resulted in an incomplete elution of the serum matrix. Instead of pre-mixed mobile phases, we used two different pumps (2A and 2B) for the delivery of the analytical mobile phase. This might appear complicated at the first view, but helped to simplify changes in the method, especially in the composition of the two components. 3.7. Analysis of patient samples Fig. 4 displays the course of posaconazole concentrations in a patient under chemotherapy for follicular lymphoma who had already received an autologous haematopoietic stem cell trans-
Fig. 4. Time course of serum concentrations of posaconazole in a patient at the end of treatment (for details see text).
Fig. 5. Posaconazole concentrations in relation to the applied doses (samples originating from five patients, n = 16).
plant. The patient came from another hospital with the unusual dosage of 400 mg posaconazole three times daily. We examined the serum concentrations over 5 days after discontinuation of the drug. The average half-life of posaconazole can be derived from the measured data to be approximately 35 h (Fig. 4). Fig. 5 shows posaconazole concentrations in relation to the applied doses (400, 600, 800 and 1200 mg daily) in five cancer patients. All blood samples (n = 16) were drawn before the intake of posaconazole as a morning trough value. The measured drug levels varied between 0.117 and 2.252 mg/l. It is obvious that there is a wide variability within the determined serum concentrations of the 800 mg/d dosage group. 4. Conclusion The presented HPLC method allows the fast and automated determination of the antifungal drug posaconazole in human serum without the need for deproteinization and thereby loss of analyte or any other form of sample pre-treatment. The first step of the analysis is an automated sample clean-up by use of an extraction column packed with a restricted access material which leads to the separation of the analyte from the serum proteins and allows the direct and repetitive injection of untreated serum samples. The second step consists of the chromatographic separation performed on an RP18 analytical column followed by identification with a fluorescence detector. Several pre-existing methods for the determination of posaconazole in human serum report slightly lower LOD or LLOQ [17,27], but the sensitivity and selectivity of our method are absolutely sufficient for routine clinical use. Furthermore, as not every hospital or laboratory has access to mass spectrometry, methods
2498
W. Neubauer et al. / J. Chromatogr. B 877 (2009) 2493–2498
based on fluorescence detection are likely to be a more suitable alternative for the standard routine. Acknowledgements We thank the Chemical Research Division, Schering-Plough Research Institute (Kenilworth, NJ, USA) for supplying the pure substance of posaconazole and the central medical chemistry laboratory of Freiburg University Medical Center for the pooled blank serum. We are also obliged to Christoph Trautwein, Armin Schuster and Andreas Längin for their help and technical assistance and to Egid Strehl for his generous support. References [1] K.A. Marr, R.A. Carter, F. Crippa, A. Wald, L. Corey, Clin. Infect. Dis 34 (2002) 909. [2] Z. Bhatti, A. Shaukat, N.G. Almyroudis, B.H. Segal, Mycopathologia 162 (2006) 1. [3] E.J. Bow, Br. J. Haematol. 140 (2008) 133. [4] M.A. Pfaller, D.J. Diekema, J. Clin. Microbiol. 42 (2004) 4419. [5] P.G. Pappas, J.H. Rex, J. Lee, R.J. Hamill, R.A. Larsen, W. Powderly, C.A. Kauffman, N. Hyslop, J.E. Mangino, S. Chapman, H.W. Horowitz, J.E. Edwards, W.E. Dismukes, Clin. Infect. Dis. 37 (2003) 634. [6] R. Herbrecht, D.W. Denning, T.F. Patterson, J.E. Bennett, R.E. Greene, J.W. Oestmann, W.V. Kern, K.A. Marr, P. Ribaud, O. Lortholary, R. Sylvester, R.H. Rubin, J.R. Wingard, P. Stark, C. Durand, D. Caillot, E. Thiel, P.H. Chandrasekar, M.R. Hodges, H.T. Schlamm, P.F. Troke, B. de Pauw, N. Engl. J. Med. 347 (2002) 408. [7] Noxafil (posaconazole), Prescribing Information, Schering Corporation, Kenilworth, NJ, October 2006 (USA). [8] Noxafil (posaconazole), European Prescribing Information, Schering-Plough Europe, Brussels, Belgium, March 2007 (European Union). [9] R. Herbrecht, Int. J. Clin. Pract. 58 (2004) 612.
[10] G.M. Keating, Drugs 65 (2005) 1553. [11] J.A. Vazquez, D.J. Skiest, L. Nieto, R. Northland, I. Sanne, J. Gogate, W. Greaves, R. Isaacs, Clin. Infect. Dis. 42 (2006) 1179. [12] R.N. Greenberg, K. Mullane, J.A. van Burik, I. Raad, M.J. Abzug, G. Anstead, R. Herbrecht, A. Langston, K.A. Marr, G. Schiller, M. Schuster, J.R. Wingard, C.E. Gonzalez, S.G. Revankar, G. Corcoran, R.J. Kryscio, R. Hare, Antimicrob. Agents Chemother. 50 (2006) 126. [13] F. Barchiesi, D. Arzeni, A.W. Fothergill, L.F. Di Francesco, F. Caselli, M.G. Rinaldi, G. Scalise, Antimicrob. Agents Chemother. 44 (2000) 226. [14] K.L. Oakley, C.B. Moore, D.W. Denning, Antimicrob. Agents Chemother. 41 (1997) 1124. [15] R. Courtney, S. Pai, M. Laughlin, J. Lim, V. Batra, Antimicrob. Agents Chemother. 47 (2003) 2788. [16] A.J. Ullmann, O.A. Cornely, A. Burchardt, R. Hachem, D.P. Kontoyiannis, K. Töpelt, R. Courtney, D. Wexler, G. Krishna, M. Martinho, G. Corcoran, I. Raad, Antimicrob. Agents Chemother. 50 (2006) 658. [17] C. Müller, M. Arndt, C. Queckenberg, O.A. Cornely, M. Theisohn, Mycoses 49 (2006) 17. [18] T.J. Walsh, I. Raad, T.F. Patterson, P. Chandrasekar, G.R. Donowitz, R. Graybill, R.E. Greene, R. Hachem, S. Hadley, R. Herbrecht, A. Langston, A. Louie, P. Ribaud, B.H. Segal, D.A. Stevens, J.A. van Burik, C.S. White, G. Corcoran, J. Gogate, G. Krishna, L. Pedicone, C. Hardalo, J.R. Perfect, Clin. Infect. Dis. 44 (2007) 2. [19] M.L. Goodwin, R.H. Drew, J. Antimicrob. Chemother. 61 (2008) 17. [20] K. Boos, C. Grimm, Trends Analyt. Chem. 18 (1999) 175. [21] R. Trittler, M. Ehrlich, T.J. Galla, R.E. Horch, K. Kümmerer, J. Chromatogr. B 775 (2002) 127. [22] M. Ehrlich, R. Trittler, F.D. Daschner, K. Kümmerer, J. Chromatogr. B 755 (2001) 373. [23] H. Egle, R. Trittler, A. König, K. Kümmerer, J. Chromatogr. B 814 (2005) 361. [24] H. Egle, R. Trittler, K. Kümmerer, Ther. Drug Monit. 26 (2004) 425. [25] H. Egle, R. Trittler, K. Kümmerer, Rapid Commun Mass Spectrom. 18 (2004) 2871. [26] S. Chhun, E. Rey, A. Tran, O. Lortholary, G. Pons, V. Jullien, J. Chromatogr. B 852 (2007) 223. [27] H. Kim, P. Likhari, P. Kumari, C.C. Lin, A.A. Nomeir, J. Chromatogr. B 738 (2000) 93. [28] J.X. Shen, G. Krishna, R.N. Hayes, J. Pharm. Biomed. Anal. 43 (2007) 228.