ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, Dec. 2010, p. 5074–5081 0066-4804/10/$12.00 doi:10.1128/AAC.00022-10 Copyright © 2010, American Society for Microbiology. All Rights Reserved.
Vol. 54, No. 12
Ultra-Performance Liquid Chromatography Mass Spectrometry and Sensitive Bioassay Methods for Quantification of Posaconazole Plasma Concentrations after Oral Dosing䌤 Bertrand Rochat,1†* Andres Pascual,2† Benoıˆt Pesse,2 Fre´de´ric Lamoth,2 Dominique Sanglard,3 Laurent A. Decosterd,4 Jacques Bille,3 and Oscar Marchetti2* Quantitative Mass Spectrometry Facility, Faculty of Biology and Medicine,1 Infectious Diseases Service, Department of Medicine,2 Division of Clinical Pharmacology, Department of Medicine,4 and Institute of Microbiology,3 Centre Hospitalier Universitaire Vaudois and University of Lausanne, 1011 Lausanne, Switzerland Received 12 January 2010/Returned for modification 29 July 2010/Accepted 26 September 2010
Posaconazole (POS) is a new antifungal agent for prevention and therapy of mycoses in immunocompromised patients. Variable POS pharmacokinetics after oral dosing may influence efficacy: a trough threshold of 0.5 g/ml has been recently proposed. Measurement of POS plasma concentrations by complex chromatographic techniques may thus contribute to optimize prevention and management of life-threatening infections. No microbiological analytical method is available. The objective of this study was to develop and validate a new simplified ultra-performance liquid chromatography tandem mass spectrometry (UPLC-MS/MS) method and a sensitive bioassay for quantification of POS over the clinical plasma concentration range. The UPLC-MS/MS equipment consisted of a triple quadrupole mass spectrometer, an electrospray ionization (ESI) source, and a C18 analytical column. The Candida albicans POS-hypersusceptible mutant (MIC of 0.002 g/ml) ⌬cdr1 ⌬cdr2 ⌬flu ⌬mdr1 ⌬can constructed by targeted deletion of multidrug efflux transporters and calcineurin genes was used for the bioassay. POS was extracted from plasma by protein precipitation with acetonitrile-methanol (75%/25%, vol/vol). Reproducible standard curves were obtained over the range 0.014 to 12 (UPLC-MS/MS) and 0.028 to 12 g/ml (bioassay). Intra- and interrun accuracy levels were 106% ⴞ 2% and 103% ⴞ 4% for UPLC-MS/MS and 102% ⴞ 8% and 104% ⴞ 1% for bioassay, respectively. The intra- and interrun coefficients of variation were 7% ⴞ 4% and 7% ⴞ 3% for UPLC-MS/MS and 5% ⴞ 3% and 4% ⴞ 2% for bioassay, respectively. An excellent correlation between POS plasma concentrations measured by UPLC-MS/MS and bioassay was found (concordance, 0.96). In 26 hemato-oncological patients receiving oral POS, 27/69 (39%) trough plasma concentrations were lower than 0.5 g/ml. The UPLC-MS/MS method and sensitive bioassay offer alternative tools for accurate and precise quantification of the plasma concentrations in patients receiving oral posaconazole.
disease (19, 20, 21). A large interindividual variability in POS pharmacokinetics has been reported (9, 18, 19, 36). A recent clinical trial has suggested that this variability influences efficacy in patients with invasive mycoses: POS plasma concentrations below 0.5 to 1 g/ml have been associated with failure of salvage therapy for refractory infections (38). A POS trough threshold of 0.5 g/ml, the MIC90 of Aspergillus and Candida spp., has been suggested for prophylaxis (1, 19, 21). An average POS concentration above 0.7 g/ml is recommended in the FDA briefing document (13). Individualized oral drug dosing guided by monitoring of POS plasma concentrations might thus contribute to improve drug efficacy (14, 21, 34). High-performance liquid chromatography with UV detection (HPLC-UV) and liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS) are the reference methods for measurement of POS plasma concentrations (1, 4, 15, 16, 17, 25, 28, 33, 37). Compared with HPLC-UV, LC-MS/MS displays a short analytical time, good sensitivity (30- to 150-fold smaller drug amounts detected on column), and high specificity. A validated LC-MS/MS method has been reported by Shen et al. using a plasma extraction followed by evaporation, an atmospheric pressure chemical ionization (APCI) source transition with one product ion, and a calibration curve ranging from 0.005 to 5 g/ml. Ultra-performance LC-MS/MS (UPLC-
Posaconazole (POS) is a new extended-spectrum azole antifungal agent with activity against Candida, Aspergillus, and emerging molds, such as zygomycetes. POS is efficacious for prevention of invasive mycoses in hematological patients, for therapy of oropharyngeal and esophageal candidiasis in immunocompromised patients, and for salvage therapy of invasive infections in patients intolerant of or refractory to standard antifungal therapy (7, 26, 35, 38). POS, a lipophilic compound, is available as an oral suspension, and its absorption is saturable and dependent on frequency of dosing and on food and fat intake (11). Oral bioavailability is decreased in patients receiving proton pump inhibitors and in those with diarrhea due to chemotherapy-induced mucositis or acute graft-versus-host
* Corresponding author. Mailing address for B. Rochat: Quantitative Mass Spectrometry Facility, Faculty of Biology and Medicine, Centre Hospitalier Universitaire Vaudois, Rue du Bugnon 46, CH1011 Lausanne, Switzerland. Phone: 41 21 314 41 58. Fax: 41 21 314 42 88. E-mail:
[email protected]. Mailing address for O. Marchetti: Infectious Diseases Service, Department of Medicine, Centre Hospitalier Universitaire Vaudois, Rue du Bugnon 46, CH-1011 Lausanne, Switzerland. Phone: 41 21 314 10 10. Fax: 41 21 314 10 18. E-mail:
[email protected]. † B.R. and A.P. contributed equally to this work. 䌤 Published ahead of print on 4 October 2010. 5074
VOL. 54, 2010
UPLC-MS AND BIOASSAYS FOR POSACONAZOLE QUANTIFICATION
FIG. 1. Chemical structures of posaconazole (top, C37H42F2N8O4; monoisotopic mass, 700.3 Da) and ketoconazole, used as an internal standard (bottom, C26H28Cl2N4O4; monoisotopic mass, 530.1 Da). Fragmentation of POS in the collision cell of the triple quadrupole produces m/z at 683.3, corresponding to a loss of water, and m/z at 127.0, corresponding to a more specific product ion. Precursor ion at m/z 701.4 ([M⫹H]⫹) with product ions at m/z 683.3 and 127.0 were selected as ion transitions for their sensitivity and specificity with CE of 32 and 60 eV, respectively.
MS/MS) is a new technique, which may further improve these performances by shortening the analytical time (10). Bioassay, a simple and inexpensive culture technique, may represent a valid alternative to chromatographic methods for quantification of azole antifungals in institutions without specialized chemistry/pharmacology labs (24, 29). However, no bioassay method is available for POS quantification in plasma. The objective of the present work was to develop, validate, and compare new UPLC-MS/MS and bioassay methods for the measurement of POS plasma concentrations over the clinical analytical range. MATERIALS AND METHODS Chemicals and reagents. POS (C37H42F2N8O4; monoisotopic mass, 700.3; purity, 100%) and the internal standard (IS) ketoconazole (C26H28Cl2N4O4; monoisotopic mass, 530.1) (Fig. 1) were kindly supplied by Schering-Plough Research Institute (Kenilworth, NJ) and purchased from Sigma-Aldrich Chemie GmbH (Steinheim, Germany), respectively. Acetonitrile (MeCN) and methanol (MeOH) were of LiChrosolv grade (Merck, Dietikon, Switzerland). Formic acid (100%) was used (Fluka, Buchs, Switzerland). Ultrapure H2O was obtained by ultrafiltration using Milli-Q UFPlus (Millipore Corp., Burlington, MA). All other chemicals were of analytical grade. For the bioassay method, tubes with lower binding capacity (LoBind; Eppendorf AG, Hamburg, Germany) were used (see “Sample preparation for UPLCMS/MS and bioassay”). Working solutions, calibration standards, and quality controls. Stock solutions of POS (1 mg/ml) and the IS (1 mg/ml) were prepared in MeOH and dimethyl sulfoxide (DMSO), respectively. The stock solution of POS was diluted in pooled citrated plasma (sodium citrate) to obtain calibration standards ([Cs] 0.014, 0.028, 0.11, 0.22, 0.70, 1.80, 2.90, 4.7, and 12.0 g/ml) and quality controls ([QCs] 0.055, 0.44, 1, and 7.5 g/ml) for both UPLC-MS/MS and bioassay methods. Cs and QC samples were stored at ⫺80°C. The IS was diluted at 0.1 g/ml in MeCN-MeOH at 75%/25% (vol/vol) which was used as a precipitation solution for the UPLC-MS/MS method. UPLC-MS/MS method. The UPLC system included a Rheos Allegro quaternary ultra-performance pump (Flux Instruments, Basel, Switzerland). The autosampler HTS-PAL (CTC PAL Analytics, Zwingen, Switzerland) maintained injection vials at 12°C. The chromatographic system was coupled with a triplestage quadrupole mass spectrometer, a TSQ Quantum discovery (Thermo Fisher, San Jose, CA), equipped with an electrospray ionization (ESI) source. Chromatography was performed with a BEH C18 analytical column (2.1 mm, 30 mm, and 1.9 m as inner diameter, length, and particle size, respectively; Waters, Milford, MA). The column was placed in an oven set at 60°C allowing a
5075
significant decrease of the column back pressure. Mobile phase A was composed of 10 mM ammonium formate in H2O containing 0.1% formic acid, mobile phase B was composed of 1% acetic acid in MeOH, and mobile phase C was composed of 1% acetic acid in MeCN. The mobile phase was delivered at 0.8 ml/min using the following stepwise gradient: 0 to 0.9 min in 90% A and 10% B, 0.9 to 1.9 min in 5% A and 95% B, 1.9 to 2 min in 100% B, 2 to 3.6 min in 100% C, 3.6 to 3.7 min in 90% A and 10% B, and 3.7 to 5.0 min in 90% A and 10% B. Total run time was 5 min. This mobile phase and gradient resulted in efficient ionization of POS and the IS and symmetric chromatographic peak shapes. Pressure ranged between 2 ⫻ 107 and 6 ⫻ 107 Pa. Retention times of POS and the IS were 1.4 and 1.5 min, respectively. The ESI source was operated in positive mode under the following conditions: capillary temperature of 350°C, ESI spray voltage of 4 kV, in-source induced dissociation voltage of 10 V, tube lens voltage of 120 V, and sheath and auxiliary gas (nitrogen) flow rates of 60 and 30 arbitrary units, respectively. The first (Q1) and third (Q3) quadrupoles were set at a 1-AMU mass resolution. The gas (argon) pressure in the collision cell (Q2) was 1 mtorr (0.13 Pa). Scan time and scan width were 0.1 s and 0.5 AMU, respectively. Unsmoothed chromatographic peaks resulted from at least 10 MS scans. For POS quantification, compounds were detected via selected reaction monitoring (SRM) in centroid mode employing the ion transition of [M⫹H]⫹ precursor ions to product ions. The selected ion transitions were at m/z 701.4 3 683.3 ⫹ 127.0 and 531.2 3 489.0 with collision energies ([CE] collision-induced dissociation) of 32/60 and 32 eV with the tube lens at 190 and 145 V for POS and the IS, respectively. Two product ions were selected for POS detection: (i) the first, obtained by loss of water, a frequent fragmentation pathway (m/z 683), gave a signal intensity that was two times higher but was poorly specific; (ii) the second (m/z 127) was more specific for POS identity but gave a signal that was two times less intense. Both ion transitions could be displayed on two different chromatograms and used to confirm the identity of POS if the presence of coeluting drugs or metabolites with the same parent ion mass is suspected. UPLC-MS/MS instrument monitoring, chromatographic data acquisition, peak integration, and quantification were performed using the Xcalibur software package (Thermo Fisher, San Jose, CA). For POS quantification, chromatographic peaks were smoothed using the boxcar algorithm set at 7. Bioassay method. (i) Candida albicans strain. The C. albicans mutant DSY2621 was constructed by targeted deletions of genes encoding membrane efflux transporters (cdr1, cdr2, flu, and mdr1) and calcineurin (cna); the MIC of POS for this strain was 0.002 g/ml (30). The strain was maintained at 4°C on Sabouraud dextrose agar plates (Difco Laboratories, Basel, Switzerland). (ii) Inoculum. A single CFU of C. albicans mutant DSY2621 was grown overnight (at 200 rpm and temperature 35°C) in 2 ml of YEPD medium (10 g/liter Bacto peptone [Difco Laboratories, Basel, Switzerland], 5 g/liter yeast extract [Difco Laboratories, Basel, Switzerland], and 20 g/liter glucose [Fluka, Steinheim, Switzerland]). An inoculum of 1.5 ⫻ 107 CFU/ml was prepared by diluting the overnight culture with 0.9% NaCl to an optical density of 0.5 arbitrary units of absorbance at a 360-nm wavelength (Novaspec II spectrophotometer; Pharmacia Biotech, Cambridge, England). The viable counts of the inoculum were verified with subcultures on Sabouraud dextrose agar plates. (iii) Medium. A total of 412.5 ml of the broth medium containing 25 g/liter agar (Becton Dickinson, Sparks, Maryland) and 8 g/liter of yeast nitrogen base (Becton Dickinson, Sparks, Maryland) was autoclaved and allowed to equilibrate at a temperature of 50°C; 50 ml of a 20% glucose solution (Sigma-Aldrich Chemie, Steinheim, Germany) was added. (iv) Agar plate. A 12.8-ml buffer solution (7.15 g of KH2PO4 [Sigma-Aldrich Chemie, Steinheim, Germany] and 50 g of Na2HPO4 䡠 2H2O [Merck, Darmstadt, Germany] in a volume of 500 ml of bidistilled H2O with a final pH of 5.5), 115 ml of broth medium (see above), and 2.6 ml of inoculum (see above) were mixed at a temperature of 50°C. The final volume of 130.4 ml was poured in a square glass plate (220 by 220 mm). The agar was allowed to solidify at room temperature. Thereafter, 24 round wells (diameter, 4 mm; capacity, 25 l) were cut using a sterile cork borer over a standard template. Each well was filled with 25 l of plasma, which was allowed to diffuse through the agar at 4°C for 2 h. The plate was then incubated at 37°C for 15 h. (v) Analysis. For the validation procedure an analytical run included one blank, eight standards, and four quality controls. Growth inhibition around wells was quantified by measuring the horizontal, the vertical, and the two diagonal diameters to the nearest 0.1 mm with a vernier caliper (see “Calibration curves”). Sample preparation for UPLC-MS/MS and bioassay. Citrated plasma samples (Cs, QCs, and unknowns) were thawed at room temperature (RT), vortex mixed, and sonicated for 5 min. Samples were vortex mixed again, and 50 l (UPLCMS/MS) or 60 l (bioassay) of plasma was transferred in 1.5-ml Eppendorf tubes. A total of 150 l (UPLC-MS/MS) or 187.5 l (bioassay) of the precipi-
5076
ROCHAT ET AL.
tation solution (MeCN-MeOH at 75%/25% [vol/vol] containing 0.1 g/ml IS only for the UPLC-MS/MS method, with a final IS concentration of 0.075 g/ml) was added in each tube. Samples were vortex mixed, sonicated for 5 min, and centrifuged at 20,000 ⫻ g at RT for 10 min. For the UPLC-MS/MS method, 160 l of supernatant was transferred into injection vials prior to analysis. For the bioassay, 200 l of supernatant was transferred into LoBind Eppendorf tubes and evaporated using a SpeedVac system (for 90 min at 60°C). Samples were reconstituted in 60 l of H2O, vortex mixed, and sonicated for 5 min; 25 l was put in a well of the agar plate. Validation procedure. The methods were validated according to the international recommendations for analytical method validation (12, 32). The validation procedure included the steps described below. (i) Limits of detection and quantification. The limit of detection was defined as the lowest concentration producing a peak distinguishable from the background noise (UPLC-MS/MS) or a zone of growth inhibition (bioassay). The lower and upper limits of quantification (LLOQ and ULOQ, respectively) were defined as the lowest and highest concentrations which can be measured with an accuracy within 80 to 120% and 85 to 115% of the nominal value, respectively, and a precision of ⫾20% and ⫾15%, respectively. (ii) Recovery and matrix effect. Extraction yield and, exclusively for the UPLCMS/MS method, matrix effect (ion suppression), and overall method recovery were evaluated with the four QC concentrations. As recommended in validation guidelines or described in previous articles (5, 12, 27, 32), three procedures (A, B, and C) were performed as follows: for procedure A, analytes were spiked in water, processed as described in the paragraph on sample preparation, and directly analyzed by UPLC-MS/MS or bioassay; for procedure B (performed exclusively for the UPLC-MS/MS method), analytes were spiked in extracted plasma samples and injected in the UPLC-MS/MS system; for procedure C, analytes were spiked in plasma samples. The complete extraction procedure was carried through, and samples were analyzed by UPLC-MS/MS system or bioassay. The means of chromatographic peak areas or growth inhibition diameters obtained after the above procedures were used for calculating ratios of these values for C/B (for UPLC-MS/MS only), B/A (for UPLC-MS/MS only), and C/A (for both UPLC-MS/MS and bioassay) for determination of the yield of extraction, the matrix effect, and the overall method recovery, respectively. Assessment of conditions potentially modifying POS quantification by interfering with its ionization. (i) Interindividual plasma variability. The effect of the origin of plasma on POS quantification was studied in plasma from eight healthy volunteers. Plasma samples were spiked with the four QC concentrations. Accuracy and precision of POS concentrations were determined by UPLC-MS/MS and bioassay. (ii) Interference of anticoagulants. The effect of different types of anticoagulants (plasma with sodium citrate, potassium EDTA, or lithium heparin) and of the absence of anticoagulant (serum) on POS quantification was studied in duplicate in plasma samples collected from healthy donors using 5-ml polypropylene tubes (Monovettes; Sarstedt, Nu ¨mbrecht, Germany). After centrifugation, samples were spiked with the four QC concentrations. Accuracy and precision of POS concentrations measured by UPLC-MS/MS and bioassay were calculated (see “Intra- and interrun accuracy and precision”). (iii) Interference of drugs. The interference of drugs that are frequently used in patients with invasive fungal infections with POS quantification by UPLCMS/MS and bioassay was evaluated. Twenty-eight compounds were studied: (a) antibacterial agents, including amoxicillin-clavulanic acid, piperacillin-tazobactam, cefepime, ceftazidime, ceftriaxone, imipenem, meropenem, clarithromycin, amikacin, metronidazole, vancomycin, trimethoprim, clindamycin, and rifampin; (b) antifungal agents, including fluconazole, itraconazole, voriconazole, caspofungin, amphotericin B, and flucytosine; (c) antiretroviral agents, including indinavir, amprenavir, saquinavir, ritonavir, efavirenz, and nelfinavir; and (d) immunosuppressive agents including cyclosporine and tacrolimus. Plasma samples were spiked with 1 g/ml of POS and of one of the above drugs, a concentration that represents the in vivo observed exposure to both compounds. POS quantifications by the UPLC-MS/MS and bioassay methods in the absence and presence of a second drug were compared. Calibration curves. For both methods, UPLC-MS/MS and bioassay, six calibration curves obtained over a 4-month period were used for the determination of the best fit over the concentration spectrum ranging between the LLOQ and ULOQ. The equation showing the lowest and most constant percentage of total bias from nominal Cs values was considered for the choice of the best-fitting model. For UPLC-MS/MS, the internal standard method was used for quantification of plasma concentrations: POS/IS ratios were plotted against nominal concentrations. For the bioassay, the external standard method was used: growth inhibition diameters were plotted against spiked concentrations. Cs were measured by the UPLC-MS/MS method (in duplicate) and bioassay (single experi-
ANTIMICROB. AGENTS CHEMOTHER. ment). For both methods, back-calculated concentrations of Cs had to be within 85% to 115% of spiked (nominal) concentrations. Intra- and interrun accuracy and precision. Accuracy (measured value/nominal value ⫻ 100) and precision (coefficient of variation; standard deviation of measured values/mean measured values ⫻ 100) of UPLC-MS/MS and bioassay were determined for the four QC solutions spiked in a single lot of citrated human plasma. For the intrarun assay (within-day), six replicates of each QC were processed in the same experiment. For the interrun assay (between-day), each QC was processed in duplicate on six different days over a 2-month period. According to international standards, accuracy within the range of 85 to 115% of nominal values and precision with a coefficient of variation of ⫾15% were required. Interobserver variability of POS quantification by bioassay. For the bioassay method, interobserver variability was assessed by comparing the measurements of POS plasma concentrations by five different laboratory technicians. The accuracy and precision of the interobserver measurements for the four QC concentrations were calculated (n ⫽ 20). POS quantification over time under different storage conditions. POS quantification with UPLC-MS/MS and bioassay over time (percent deviation of measured concentrations from baseline values) was studied at different temperatures and in different biomatrices using freshly prepared Cs. Four different storage conditions (A to D) were studied: for condition A, QC samples were stored at room temperature, 4°C, and ⫺80°C; for B, citrated whole blood from a patient receiving POS therapy was stored at room temperature and 4°C and centrifuged just before extraction; for C, citrated plasma samples obtained from patients receiving POS therapy were stored at ⫺80°C; and for D, QC samples were analyzed during four freeze-thaw cycles. Correlation of POS concentrations measured by UPLC-MS/MS and bioassay. The correlation of the POS concentrations measured by UPLC-MS/MS and bioassay in plasma samples from patients treated with POS was analyzed according to Lin (23), as follows (where is standard deviation of measured values and is the mean of measured values): scale shift, u ⫽ (1 ⫺ 2)/公(1 ⫺ 2); location shift, v ⫽ 1/2; precision, r (the coefficient of correlation); congruence, A ⫽ [(v ⫹ 1/v ⫹ u2)] ⫺ 1; and concordance, rA. POS plasma concentrations in clinical samples. After conclusion of the validation procedure, POS plasma trough concentrations were measured in hematooncological patients receiving oral prophylaxis against or therapy for invasive mycoses. These measurements were performed at the request of the treating physicians in the context of the anti-infective therapeutic drug monitoring program at our institution. Blood samples (5 ml) were collected in tubes containing sodium citrate 5 min before the administration of the subsequent dose. Tubes were centrifuged at 4°C within 1 h and stored at ⫺80°C until analysis by UPLCMS/MS. Data on patient demographics, indications for POS, and trough plasma concentrations were analyzed. The proportion of measurements below 0.5 g/ml, the efficacy threshold proposed in recent studies, was calculated (1, 3, 19, 21, 22, 38).
RESULTS UPLC-MS/MS chromatograms after injections of extracted plasma samples spiked with POS at 0.014 g/ml (LLOQ) and of patient plasma extract are shown in Fig. 2A and B, respectively. The chromatograms of extracted blank plasma injected after plasma spiked with 12 g/ml POS (ULOQ) showed the absence of carryover (data not shown). Adsorption of POS. During the initial development of the two methods, an inconsistent reproducibility of the quantification results was observed. This phenomenon could be attributed to POS adsorption on plastic, glassware, and/or plasma constituents (data not shown). The inclusion of sonication and centrifugation at RT in the extraction procedure prevented POS adsorption and substantially improved robustness and reproducibility of quantification. Recoveries and matrix effect. The extraction yield (UPLCMS/MS and bioassay), matrix effect, and overall method recovery (UPLC-MS/MS) of POS were independent from the concentrations: mean values of the four QCs are shown in the Table 1. The extraction yield, matrix effect, and total method
VOL. 54, 2010
UPLC-MS AND BIOASSAYS FOR POSACONAZOLE QUANTIFICATION
FIG. 2. UPLC-MS/MS chromatograms of extracted samples with ion transitions for posaconazole (POSA) and IS. (A) Chromatogram after injection (6 l) of a calibrator sample (Cs) spiked with 0.014 g/ml POS (equivalent to the LLOQ) and 0.075 g/ml IS. (B) Chromatogram after injection of a plasma extract from a patient receiving POS therapy. The specific ion transitions (m/z; SRM) are reported. NL, signal intensity in arbitrary units; RT, retention time in min; AA, peak area in arbitrary units; ⫹ c ESI, electrospray ionization in centroid and positive mode of acquisition.
5077
recovery of ketoconazole (the IS for the UPLC-MS/MS method) were 110% ⫾ 18, 89% ⫾ 9, and 97% ⫾ 18%, respectively. Assessment of conditions potentially modifying POS quantification by interfering with its ionization. (i) Interindividual plasma variability. The mean coefficients of variation of POS quantification in the four QC samples spiked in plasma from eight healthy volunteers were 3.9% (n ⫽ 32) and 1.5% (n ⫽ 24) for UPLC-MS/MS and bioassay, respectively. These results suggest that the analytical procedure is unlikely to be modified by interindividual variations of the biomatrix. (ii) Interferences of anticoagulants. Using POS Cs samples spiked in citrated plasma, measurements of POS concentrations in the four QC samples spiked in potassium EDTA, in lithium heparin plasma, or in serum were accurate for the bioassay (mean measured values within 85 to 115% of nominal concentrations) but inaccurate for the UPLC-MS/MS method (mean measured values of 150% ⫾ 11%, 137% ⫾ 16%, and 165% ⫾ 11% of nominal concentrations, respectively). In contrast, quantification of POS by UPLC-MS/MS was accurate if the same anticoagulant was used for Cs and QCs (mean of measured values within 85 to 115% of nominal concentrations for potassium EDTA, lithium heparin plasma, or serum). (iii) Interferences of drugs. With the 28 tested compounds (see Materials and Methods), the accuracy of POS quantification by UPLC-MS was within 85 to 115% of the nominal concentrations. Whereas no interference with POS quantification by bioassay was observed in the presence of 26 compounds, the method overestimated the nominal POS concentrations by a factor 2.1 and 2.9 in the presence of voriconazole and flucytosine, respectively. Calibration curves and lower and upper limits of quantification. The LLOQ, 0.014 for UPLC-MS/MS and 0.028 g/ml for bioassay (Table 1), correspond to the detection of 14 pg on the column and of 560 pg in the well, respectively. For both methods, the ULOQ were 12 g/ml (Table 1). The calibration curves of UPLC-MS/MS and bioassay included nine and eight standards, respectively. Taking into account POS plasma concentrations in patients receiving recommended doses, both methods cover the clinical analytical range (8, 19, 31). For the UPLC-MS/MS method, a quadratic equation weighted by the inverse-squared concentration (1/x2) resulted in the best fit (Fig. 3) with a curve regression coefficient of ⬎0.99. For the bioassay, a cubic equation (regression order 3) resulted in the best fit (Fig. 4) with a curve regression coefficient of ⬎0.99. Intra- and interrun accuracy and precision. The means ⫾ standard deviations of accuracy and precision of intra- and interrun assays for the four QCs are reported in Table 1 for both methods. Mean interobserver (five different laboratory technicians) accuracy and precision of POS quantification by bioassay were 103% ⫾ 5% and 5% ⫾ 3%, respectively. POS stability over time under different storage conditions. POS quantifications in spiked samples and in patient samples were stable over time for the tested storage conditions (measured concentrations within 85 to 115% of the values at time zero) (Table 1). Correlation of POS quantifications by UPLC-MS/MS and bioassay. The results of the POS concentrations measured in 25 clinical plasma samples by UPLC-MS/MS and bioassay are
5078
ROCHAT ET AL.
ANTIMICROB. AGENTS CHEMOTHER. TABLE 1. Validation data of UPLC-MS/MS and bioassay methods Value for the parameter by:
Validation experiment parametera
UPLC-MS/MS
LLOQ (g/ml) Accuracy (% 关n ⫽ 6兴) Precision (% 关n ⫽ 6兴)
Bioassay
0.014 100.9 ⫾ 7.6 7.5
0.028 100.5 ⫾ 0.9 2.4
0.014–12
0.028–12
Analytical range (g/ml)* Intraday assay (% 关n ⫽ 5兴)* Accuracy Precision
106 ⫾ 2 7⫾4
102 ⫾ 8 5⫾3
Interday assay (% 关n ⫽ 5兴)* Accuracy Precision
103 ⫾ 4 7⫾3
104 ⫾ 1 4⫾2
Duration of POS quantification values within ⫾15% of values measured at time zero, by condition At 4°C At room temp At ⫺80°C No. of freeze-thaw cycles In whole blood At 4°C At room temp Extraction yield (%)* Matrix effect (%)* Total recovery (%)* Interference of the interindividual variability of the biological matrix with POS quantification (plasma from 8 different donors)* Interference of drugs with POS quantification (28 compounds)* Interference of the anticoagulant in the biological matrix with POS quantification*b
ⱖ4 days ⱖ4 days ⱖ6 months ⱖ4 cycles
ⱖ4 days ⱖ4 days ⱖ6 months ⱖ4 cycles
ⱖ4 days ⱖ4 days
ⱖ4 days ⱖ4 days
82 ⫾ 11 88 ⫾ 18 70 ⫾ 6
84 ⫾ 4 NAc NA No difference
No difference No interference Interference with POS quantification: the same anticoagulant in the biological matrix is required for Cs, QCs, and patients samples
Voriconazole and flucytosine increase the quantification of POS POS quantification independent of presence/absence and type of anticoagulant
Results for the four QCs (indicated by *) are reported as mean values ⫾ standard deviations. Anticoagulants were plasma with sodium citrate, potassium EDTA, and lithium heparin; the absence of anticoagulant was studied using serum. c NA, not applicable. a b
shown in Fig. 5. The correlation coefficient was 0.97 (P ⬍ 0.001), the congruence was 0.99, and the concordance was 0.96. POS plasma concentrations in patients receiving oral prophylaxis or therapy. POS trough plasma concentrations were measured at the request of the treating physicians in 26 patients (62% male and 38% female; median age, 53 years; age range, 10 to 72 years) with hematological malignancies. Figure 6 shows the important variability of POS trough plasma concentrations measured by UPLC-MS/MS in 69 samples (median, 2.5 measurements per patient; range, 1 to 6); the median was 0.67 g/ml, with a range of 0.02 to 4.71 g/ml. As illustrated in Fig. 6, a large intra- and interindividual variability of POS trough plasma concentrations has been observed at the recommended oral dosing schedules. Seven of 26 (27%) POS trough concentrations in patients receiving the prophylactic dosing regimen (i.e., 600 mg/day) were below 0.5 g/ml. Twenty of 43 (46.5%) POS trough concentrations in those receiving the therapeutic dosing regimen (i.e., 800 mg/day) were below 0.5 g/ml (no differences among troughs obtained
with the dosing schedules of 200 mg four times a day [q.i.d.] and 400 mg twice a day [b.i.d.] were observed).
DISCUSSION High-quality validated analytical tools are needed for monitoring POS blood concentrations, which might help to improve efficacy of prevention and therapy in patients at high risk of or with life-threatening invasive fungal infections (2). Robust, accurate, and precise UPLC-MS/MS and sensitive bioassay methods for quantification of POS in plasma over the clinical analytical range were developed and validated according to international guidelines. Data for plasma extraction by protein precipitation, run time, and amount of POS detected on-column from the present and a previously reported LC-MS/MS technique were similar (33). However, the sample extraction without evaporation, the use of an ESI source, the determination of POS with two product ions (one more sensitive and one more specific), and the ex-
VOL. 54, 2010
UPLC-MS AND BIOASSAYS FOR POSACONAZOLE QUANTIFICATION
FIG. 3. UPLC-MS/MS POS calibration curve in plasma. POS concentrations are plotted against ratios of POS/IS chromatographic peak areas. The best fit is obtained using a quadratic regression and a weighting by the inverse-squared values (1/x2). The low POS concentration range is zoomed and depicted in the inset panel at the top left. The variability (standard deviation) of the IS chromatographic peak areas of Cs, QCs, and patient samples is compared with that of blank plasma samples and water in the inset panel at the bottom right.
5079
FIG. 5. Correlation of POS concentrations measured in 25 clinical plasma samples by UPLC-MS/MS and bioassay using C. albicans ⌬cdr1 ⌬cdr2 ⌬flu ⌬mdr1 ⌬cna. Scale shift, u, is ⫺0.0069; location shift, v, is 1.283; coefficient of correlation, r, is 0.97 (P ⬍ 0.001); congruence, A, is 0.99; and concordance, rA, is 0.96.
tended analytical range (0.014 to 12 g/ml) characterize the robustness and clinical utility of the new UPLC-MS/MS method. The LLOQ could be further lowered at least 10 times if a larger plasma volume (e.g., 500 l) and an evaporation reconstitution step were used, which may be relevant for research purposes in experimental models of infection. The sensitivity, accuracy, and precision of the first reported bioassay method using a POS-hypersusceptible C. albicans mutant (POS MIC of 0.002 g/ml) are comparable to data from the UPLC-MS/MS. As POS is highly bound to plasma proteins, the bioassay required an extraction by protein precipitation, as with the UPLC-MS/MS method. The excellent correlation of the POS concentrations measured by both methods in the
cross-validation studies with clinical samples confirmed the equivalent analytical performance of the bioassay. Testing of POS quantification over time under different storage conditions, using plasma from different sources containing different types of anticoagulants, and in the presence of frequently used comedications provided useful additional data for improving the bedside procedures, the preanalytical handling of biological samples, and the reliability of the interpretation of the measured POS plasma concentrations. The absence of significant changes in POS quantification over time at room temperature, at 4°C, and at ⫺80°C suggests that the compound is stable under real-life conditions, i.e., that no significant preanalytical back transformation of hypothetical unstable metabolites to the parent drug and/or degradation of POS occurs in vitro. However, no stability testing was performed at ⫺20°C, which may be relevant for implementing the method in nonresearch labs. Of note, the use of the same anticoagulant for calibrators and patient samples is mandatory for the UPLC-
FIG. 4. Bioassay calibration curve in plasma. Growth inhibition diameters are plotted against POS concentrations. The best fit is obtained with a cubic regression.
FIG. 6. POS trough plasma concentrations in patients on a prophylactic oral dosing schedule (200 mg three times a day [t.i.d.], 26 measurements;) or a therapeutic oral dosing schedule (200 mg q.i.d., 32 measurements, or 400 mg b.i.d., 11 measurements;). Each symbol represents one single measurement. Horizontal bars indicate median values for each dosing schedule.
5080
ROCHAT ET AL.
MS/MS method, whereas the type of anticoagulant does not influence the quantification by bioassay. No interference of comedications with POS quantification by UPLC-MS/MS was observed, while the presence of flucytosine, an antifungal agent which may occasionally be used in combination with POS, results in an overestimation of POS plasma concentrations by bioassay. An important variability of trough plasma concentrations was observed in hematological patients receiving oral POS for prophylaxis or therapy. These data corroborate previous observations supporting therapeutic drug monitoring for optimization of the efficacy of oral POS dosing in patients at risk of or with invasive fungal infections (1, 3, 19, 21, 22, 38). However, it is unclear whether plasma concentrations reliably reflect drug penetration and antifungal activity at the site(s) of infection: the clinical significance of POS plasma concentrations below the proposed trough threshold of 0.5 g/ml thus needs to be confirmed in prospective investigations (6). A limitation of the present methods is the quantification of a single antifungal agent. While the simultaneous quantification of multiple compounds by a single LC-MS/MS analytical run has been reported (4, 15), this is not feasible with a bioassay. This additional aspect was not considered since the primary objective of this analytical development was the validation and comparison of two alternative methods for the accurate and precise quantification of POS in lab settings with or without specialized chromatographic equipments. Thus, it remains to be investigated whether an LC-MS/MS method with identical or similar experimental conditions could be simultaneously used for different antifungal agents. In conclusion, the UPLC-MS/MS technology, which is considered the analytical gold standard for drug quantification and therapeutic drug monitoring, and the traditional microbiological method are accurate and precise for POS quantification in human plasma. These two methods may be complementarily implemented for clinical and research purposes in different laboratory settings. Expensive equipment handled by specialized technical personnel makes this rapid UPLC-MS/MS method suitable and cost-effective for reference laboratories centralizing a high number of analyses and providing results within 12 to 24 h. Cheap and simple-to-use instrumentation handled by unspecialized technical personnel makes this sensitive bioassay an excellent alternative for analyses of small numbers of samples in microbiology laboratories of institutions without specialized chemistry/pharmacology labs.
ANTIMICROB. AGENTS CHEMOTHER.
3. 4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
ACKNOWLEDGMENTS We are thankful to Schering-Plough Laboratories, Luzern, Switzerland, and to the Foundation for the Advancement in Medical Microbiology and Infectious Diseases, FAMMID, Lausanne, Switzerland, for the unrestricted grant support, and the Loterie Romande, Lausanne, Switzerland, for the support in the acquisition of the TSQ mass spectrometer. We thank Veronica Dubbini for the help in the manuscript preparation.
19.
20.
21.
REFERENCES 1. Andes, D., A. Pascual, and O. Marchetti. 2009. Antifungal therapeutic drug monitoring: established and emerging indications. Antimicrob. Agents Chemother. 53:24–34. 2. Bru ¨ggemann, R. J., D. J. Touw, R. E. Aarnoutse, P. E. Verweij, and D. M. Burger. 2009. International interlaboratory proficiency testing program for
22. 23. 24.
measurement of azole antifungal plasma concentrations. Antimicrob. Agents Chemother. 53:303–305. Bryant, A. M., D. Slain, A. Cumpston, and M. Craig. 2009. Abstr. 49th Intersci. Conf. Antimicrob. Agents Chemother., abstr. A1-590. Chhun, S., E. Rey, A. Tran, O. Lortholary, G. Pons, and V. Jullien. 2007. Simultaneous quantification of voriconazole and posaconazole in human plasma by high-performance liquid chromatography with ultra-violet detection. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 852:223–228. Colombo, S., A. Beguin, A. Telenti, J. Biollaz, T. Buclin, B. Rochat, and L. A. Decosterd. 2005. Intracellular measurements of anti-HIV drugs indinavir, amprenavir, saquinavir, ritonavir, nelfinavir, lopinavir, atazanavir, efavirenz and nevirapine in peripheral blood mononuclear cells by liquid chromatography coupled to tandem mass spectrometry. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 819:259–276. Conte, J. E., Jr., J. A. Golden, G. Krishna, M. McIver, E. Little, and E. Zurlinden. 2009. Intrapulmonary pharmacokinetics and pharmacodynamics of posaconazole at steady state in healthy subjects. Antimicrob. Agents Chemother. 53:703–707. Cornely, O. A., J. Maertens, D. J. Winston, J. Perfect, A. J. Ullmann, T. J. Walsh, D. Helfgott, J. Holowiecki, D. Stockelberg, Y. T. Goh, M. Petrini, C. Hardalo, R. Suresh, and D. Angulo-Gonzalez. 2007. Posaconazole vs. fluconazole or itraconazole prophylaxis in patients with neutropenia. N. Engl. J. Med. 356:348–359. Courtney, R., A. Sansone, W. Smith, T. Marbury, P. Statkevich, M. Martinho, M. Laughlin, and S. Swan. 2005. Posaconazole pharmacokinetics, safety, and tolerability in subjects with varying degrees of chronic renal disease. J. Clin. Pharmacol. 45:185–192. Courtney, R., D. Wexler, E. Radwanski, J. Lim, and M. Laughlin. 2004. Effect of food on the relative bioavailability of two oral formulations of posaconazole in healthy adults. Br. J. Clin. Pharmacol. 57:218–222. Cunliffe, J. M., C. F. Noren, R. N. Hayes, R. P. Clement, and J. X. Shen. 2009. A high-throughput LC-MS/MS method for the quantitation of posaconazole in human plasma: implementing fused core silica liquid chromatography. J. Pharm. Biomed. Anal. 50:46–52. Ezzet, F., D. Wexler, R. Courtney, G. Krishna, J. Lim, and M. Laughlin. 2005. Oral bioavailability of posaconazole in fasted healthy subjects: comparison between three regimens and basis for clinical dosage recommendations. Clin. Pharmacokinet. 44:211–220. Food and Drug Administration. 2001. Guidance for industry: bioanalytical method validation. Food and Drug Administration, Rockville, MD. http: //www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation /Guidances/UCM070107.pdf. Food and Drug Administration. 2009. 4 January 2010, accession date. Posaconazole. FDA briefing document. Food and Drug Administration, Rockville, MD. http://www.aapsj.org/abstracts/AM_2007/AAPS2007-000993 .PDF. Hope, W. W., and G. L. Drusano. 2009. Antifungal pharmacokinetics and pharmacodynamics: bridging from the bench to bedside. Clin. Microbiol. Infect. 15:602–612. Kahle, K., P. Langmann, D. Schirmer, U. Lenker, D. Keller, A. Helle, H. Klinker, and W. J. Heinz. 2009. Simultaneous determination of voriconazole and posaconazole concentrations in human plasma by high-performance liquid chromatography. Antimicrob. Agents Chemother. 53:3140–3142. Kim, H., C. C. Lin, M. Laughlin, R. Lovey, A. Saksena, L. Heimark, and A. A. Nomeir. 2000. Chiral high-performance liquid chromatographic analysis of antifungal SCH 56592 and evaluation of its chiral inversion in animals and humans. Chirality 12:590–597. Kim, H., P. Kumari, C. C. Lin, and A. A. Nomeir. 2002. Simultaneous high-performance liquid chromatographic determination of SCH 59884 (phosphate ester prodrug of SCH 56592), SCH 207962 and SCH 56592 in dog plasma. J. Pharm. Biomed. Anal. 27:295–303. Kohl, V., C. Mu ¨ller, O. A. Cornely, K. Abduljalil, U. Fuhr, J. J. Vehreschild, C. Scheid, M. Hallek, and M. J. Ru ¨ping. 2010. Factors influencing pharmacokinetics of prophylactic posaconazole in patients undergoing allogeneic stem cell transplantation. Antimicrob. Agents Chemother. 54:207–212. Krishna, G., M. Martinho, P. Chandrasekar, A. J. Ullmann, and H. Patino. 2007. Pharmacokinetics of oral posaconazole in allogeneic hematopoietic stem cell transplant recipients with graft-versus-host disease. Pharmacotherapy 27:1627–1636. Krishna, G., A. Moton, L. Ma, M. M. Medlock, and J. McLeod. 2009. Pharmacokinetics and absorption of posaconazole oral suspension under various gastric conditions in healthy volunteers. Antimicrob. Agents Chemother. 53:958–966. Lebeaux, D., F. Lanternier, C. Elie, F. Suarez, A. Buzyn, J. P. Viard, M. E. Bougnoux, M. Lecuit, V. Jullien, and O. Lortholary. 2009. Therapeutic drug monitoring of posaconazole: a monocentric study with 54 adults. Antimicrob. Agents Chemother. 53:5224–5229. Leventakos, K., R. E. Lewis, I. I. Raad, and D. P. Kontoyiannis. 2009. Abstr. 49th Intersci. Conf. Antimicrob. Agents Chemother., abstr. M-1054. Lin, L. I. 1989. A concordance correlation coefficient to evaluate reproducibility. Biometrics 45:255–268. Marchetti, O., P. A. Majcherczyk, M. P. Glauser, J. Bille, P. Moreillon, and
VOL. 54, 2010
25.
26.
27.
28.
29.
30.
31.
32.
UPLC-MS AND BIOASSAYS FOR POSACONAZOLE QUANTIFICATION
D. Sanglard. 2001. Sensitive bioassay for determination of fluconazole concentrations in plasma using a Candida albicans mutant hypersusceptible to azoles. Antimicrob. Agents Chemother. 45:696–700. Muller, C., M. Arndt, C. Queckenberg, O. A. Cornely, and M. Theisohn. 2006. HPLC analysis of the antifungal agent posaconazole in patients with haematological diseases. Mycoses 49(Suppl. 1):17–22. Nagappan, V., and S. Deresinski. 2007. Reviews of anti-infective agents: posaconazole: a broad-spectrum triazole antifungal agent. Clin. Infect. Dis. 45:1610–1617. Naidong, W., H. Bu, Y. L. Chen, W. Z. Shou, X. Jiang, and T. D. Halls. 2002. Simultaneous development of six LC-MS-MS methods for the determination of multiple analytes in human plasma. J. Pharm. Biomed. Anal. 28:1115– 1126. Nomeir, A. A., P. Kumari, M. J. Hilbert, S. Gupta, D. Loebenberg, A. Cacciapuoti, R. Hare, G. H. Miller, C. C. Lin, and M. N. Cayen. 2000. Pharmacokinetics of SCH 56592, a new azole broad-spectrum antifungal agent, in mice, rats, rabbits, dogs, and cynomolgus monkeys. Antimicrob. Agents Chemother. 44:727–731. Pascual, A., V. Nieth, T. Calandra, J. Bille, S. Bolay, L. A. Decosterd, T. Buclin, P. A. Majcherczyk, D. Sanglard, and O. Marchetti. 2007. Variability of voriconazole plasma levels measured by new high-performance liquid chromatography and bioassay methods. Antimicrob. Agents Chemother. 51: 137–143. Sanglard, D., F. Ischer, O. Marchetti, J. Entenza, and J. Bille. 2003. Calcineurin A of Candida albicans: involvement in antifungal tolerance, cell morphogenesis and virulence. Mol. Microbiol. 48:959–976. Sansone-Parsons, A., G. Krishna, J. Simon, P. Soni, B. Kantesaria, J. Herron, and R. Stoltz. 2007. Effects of age, gender, and race/ethnicity on the pharmacokinetics of posaconazole in healthy volunteers. Antimicrob. Agents Chemother. 51:495–502. Shah, V. P., K. K. Midha, J. W. Findlay, H. M. Hill, J. D. Hulse, I. J.
33.
34.
35.
36.
37.
38.
5081
McGilveray, G. McKay, K. J. Miller, R. N. Patnaik, M. L. Powell, A. Tonelli, C. T. Viswanathan, and A. Yacobi. 2000. Bioanalytical method validation—a revisit with a decade of progress. Pharm. Res. 17:1551–1557. Shen, J. X., G. Krishna, and R. N. Hayes. 2007. A sensitive liquid chromatography and mass spectrometry method for the determination of posaconazole in human plasma. J. Pharm. Biomed. Anal. 43:228–236. Thompson, G. R., III, M. G. Rinaldi, G. Pennick, S. A. Dorsey, T. F. Patterson, and J. S. Lewis II. 2009. Posaconazole therapeutic drug monitoring: a reference laboratory experience. Antimicrob. Agents Chemother. 53:2223– 2224. Ullmann, A. J., J. H. Lipton, D. H. Vesole, P. Chandrasekar, A. Langston, S. R. Tarantolo, H. Greinix, W. Morais de Azevedo, V. Reddy, N. Boparai, L. Pedicone, H. Patino, and S. Durrant. 2007. Posaconazole or fluconazole for prophylaxis in severe graft-versus-host disease. N. Engl. J. Med. 356:335–347. Ullmann, A. J., O. A. Cornely, A. Burchardt, R. Hachem, D. P. Kontoyiannis, K. Topelt, R. Courtney, D. Wexler, G. Krishna, M. Martinho, G. Corcoran, and I. Raad. 2006. Pharmacokinetics, safety, and efficacy of posaconazole in patients with persistent febrile neutropenia or refractory invasive fungal infection. Antimicrob. Agents Chemother. 50:658–666. Vogeser, M., C. Rieger, H. Ostermann, and U. Spo ¨hrer. 2009. A routine method for the quantification of the novel antimycotic drug posaconazole in plasma using liquid chromatography-tandem mass spectrometry. Clin. Chem. Lab. Med. 47:579–584. Walsh, T. J., 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, and J. R. Perfect. 2007. Treatment of invasive aspergillosis with posaconazole in patients who are refractory to or intolerant of conventional therapy: an externally controlled trial. Clin. Infect. Dis. 44:2–12.
