Anal. Chem. 2003, 75, 1812-1818
Direct Plasma Analysis of Drug Compounds Using Monolithic Column Liquid Chromatography and Tandem Mass Spectrometry Yunsheng Hsieh,*,† Ganfeng Wang,† Yuguang Wang,‡ Samuel Chackalamannil,‡ and Walter A. Korfmacher†
Drug Metabolism and Pharmacokinetics Department and Chemistry Department, Schering-Plough Research Institute, Kenilworth, New Jersey 07033
A monolithic silica column high-performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS) method has been developed for the high-speed direct simultaneous determination of a drug discovery compound and its major circulating metabolite (M-72) in rat plasma. This methodology makes use of flow programming and an alkyl-bonded silica rod column for fast macromolecule removal and chromatographic separation without the need for significant sample preparation. The matrix ionization suppression effect on the monolithic column HPLC-MS/MS system was investigated using the postcolumn infusion technique. After 200 plasma injections on a 50 × 4.6 mm monolithic silica column, consistent column efficiency of close to 39 000 theoretical plates/m and reproducible retention times for the analytes were observed. The apparent on-column recoveries of 12 test compounds in rat plasma samples were greater than 90%. The proposed fast direct plasma injection method was tested over a 3-day period with the interday coefficient of variation less than 15% for both analytes. In accelerating the discovery process for pharmaceutical drugs, there is a continuing demand for bioanalytical methods possessing high sensitivity and high sample throughput for both in vitro and in-vivo drug metabolism and pharmacokinetic (PK) evaluations.1,2 The high-resolution power of chromatographic methodologies coupled to atmospheric pressure ionization tandem mass spectrometry (API-MS/MS) has been able to reduce the need for most traditional sample preparation procedures and has also reduced the method development time required for drug analyses.3,4 Furthermore, fast HPLC techniques in combination with the specificity of MS/MS detection have successfully demonstrated the capability of separating and identifying a wide range of small * Corresponding author. E-mail:
[email protected]. † Drug Metabolism and Pharmacokinetics Department. ‡ Chemistry Department. (1) White R. E. Annu. Rev. Pharmacol. Toxicol. 2000, 40, 133. (2) Korfmacher, W. A.;. Cox, K. A.; Bryant, M. S.; Veals, J.; Ng, K.; Watkins, R.; Lin, C. Drug Discovery Today 1997, 2, 532. (3) Huang, E. C.; Wachs, T.; Conboy, J.; Henion, J. D. Anal. Chem. 1990, 62, 713A. (4) Miller-Stein, C.; Bonfiglio, R.; Olah, T. V.; King, R. C. Am. Pharm. Rev. 2001, 3, 54.
1812 Analytical Chemistry, Vol. 75, No. 8, April 15, 2003
molecules within 1-min gradient or isocratic analyses.5-9 However, certain sample preparation steps such as the protein precipitation procedure to remove proteins from biological samples are still essential prior to the HPLC-MS/MS assay for small molecules to prevent the HPLC column from clogging in reversed-phase chromatography and to avoid ion source contamination in the mass spectrometer.10,11 In some cases, sample preparation has become a bottleneck in the bioanalytical process. One popular HPLC packing material designed for direct assay of small molecules in biological fluids is the restricted access material (RAM),12-13 such as alkyldiol silica (ADS).14 These restricted access media permit the effective and repetitive extraction of a wide variety of compounds in untreated biological samples by preventing access of macromolecules based on a size-exclusion process. Low-molecular-mass analytes are retained by conventional retention mechanisms. Normally, the ADS column is combined with an analytical column for chromatographic separation using the column-switching technique, also called the coupled-column mode or LC-LC.15,16 Because of this sequential process, analytical throughput is limited for single-column (nonparallel) systems. An alternative methodology for direct plasma injection using a singlepolymer-coated mixed-functional (PCMF) column HPLC-MS/ MS system for simultaneously providing both protein removal and chromatographic separation (single-column mode) was developed in our laboratory in support of pharmacokinetic screening 17-20 and (5) Cheng, Y.; Lu, Z.; Neue, U. Rapid Commun. Mass Spectrom. 2001, 15, 141 (6) Rule, G.; Chapple, M.; Henion, J. D. Anal. Chem. 2001, 73, 439. (7) Hsieh, Y.; Brisson J.; Wang, G.; Ng, K.; Korfmacher, W. A. J. Pharm. Biomed. Anal., in press. (8) Hsieh, Y.; Chintala, M.; Hong, M.; Agans, J.; Brisson J.; Ng, K.; Korfmacher, W. A. Rapid Commun. Mass Spectrom. 2001, 15, 2481. (9) Hsieh, Y.; Wang, G.; Wang, Y.; Chackalamannil, S.; Brisson J.; Korfmacher, W. A. Rapid Commun. Mass Spectrom. 2002, 16, 944. (10) Henion, J.; Prosser, S. J.; Corso, T. N.; Schultz, G. A. Am. Pharm. Rev. 2000, 3, 19. (11) Henion, J.; Brewer, E.; Rule, G. Anal. Chem. 1998, 70, 650A. (12) Boos, K.; Grimm, C. Trends Anal. Chem. 1999, 18, 175. (13) Hennion, M. J. Chromatogr., A 1999, 856, 3. (14) Hogendoorn, E. A.; Zoonen, P. V.; Polettini, A.; Bouland G. M.; Montagna, M. Anal. Chem. 1998, 70, 1362. (15) Baeyens, W.; Weken, G.; Haustraete, J.; Aboul-Enein, H. Y.; Corveleyn, S.; Remon, J.; Garcia-Campana, A.; Deprez, P. J. Chromatogr., A 2000, 871, 153. (16) Schafer, C.; Lubda, D. J. Chromatogr., A 2001, 909, 73. (17) Hsieh, Y.; Bryant, M.; Gruela, G.; Brisson, J.; Ng, K.; Korfmacher, W. A. Rapid Commun. Mass Spectrom. 2000, 14, 1384. 10.1021/ac020630e CCC: $25.00
© 2003 American Chemical Society Published on Web 03/21/2003
plasma stability measurements.21,22 In this study, we explore the utility of a monolithic silica column offering the same direct plasma injection function as the PCMF column method but with at least 4 times higher sample throughput and a greater column efficiency for the separation of small molecules. This new stationary phase is obtained by the hydrolysis and polycondensation of alkoxysilanes and possesses large through-pores for macromolecules to pass through the column while retaining the drug molecules on the bonded reversed phase for chromatographic interaction. Ionization suppression (also known as matrix effects) of the analytes has been recognized as a concern in HPLC-MS/MS assays and is typically attributed to salts and coeluting nonvolatile components from the biological matrix9,23-26. For this work, matrix effects were explored using the postcolumn infusion technique.25,26 The performance of the proposed method was further evaluated using both study samples and spiked plasma samples containing clozapine and 11 drug discovery compounds. The assay accuracy was tested using three interday determinations. The analytical results obtained following direct injection of either diluted plasma or plasma extract using the protein precipitation procedure were in good agreement for both the dosed compound and its amine metabolite. EXPERIMENTAL SECTION Reagents and Chemicals. The test compound I (RNHCOOC2H5), its amine metabolite (R-NH2) compound II, and compounds III-XI were new chemical entities obtained from Schering-Plough Research Institute. The chemical structure of the internal standard (ISTD) was published elsewhere.27 Clozapine (8-chloro-11-(4-methyl-1-piperazinyl)-5H-dibenzo[b,e][1,4]-diazepine) and trifluoroacetic acid (TFA, >98%) were purchased from Sigma (St. Louis, MO). Acetonitrile (HPLC grade) was purchased from Fisher Scientific (Pittsburgh, PA). Ammonium acetate (99.999%) was purchased from Aldrich Chemical Co., Inc. (Milwaukee, WI). Deionized water was generated from a Milli-Q water purifying system purchased from Millipore Corp. (Bedford, MA), and house high-purity nitrogen (99.999%) was used. Drug-free rat plasma was purchased from Bioreclamation Inc. (Hicksville, NY). Ammonium acetate solution (1 M, pH 7.5) was prepared by dissolving 77.1 g of ammonium acetate in 1.0 L of deionized water. (18) Hsieh, Y.; Brisson, J.; Ng, K.; White, R. E.; Korfmacher, W. A. Analyst 2001, 126, 1239. (19) Hsieh, Y.; Bryant, M.; Brisson, J.; Ng, K.; Korfmacher, W. A. J. Pharm. Biomed. Anal. 2002, 27, 285. (20) Hsieh, Y.; Bryant, M.; Brisson, J.; Ng, K.; Korfmacher, W. A. J. Chromatogr., B 2002, 767, 353. (21) Wang, G.; Hsieh, Y.; Lau, J.; Cheng, K.-C.; Ng, K.; Korfmacher, W. A.; White, R. J. Chromatogr., B. 2002, 780, 451. (22) Wang, G.; Hsieh, Y.; Cheng, K.-C.; Ng, K.; Korfmacher, W. A. SpectroscopyInt. J., in press. (23) Matuszewski, B. K.; Constanzer, M. L.; Chavez-Eng, C. M. Anal. Chem. 1998, 70, 882. (24) Mei, H.; Hsieh, Y.; Nardo C.; Xu X.; Wang, S.; Ng, K.; Korfmacher, W. A. Rapid Commun. Mass Spectrom. 2003, 17, 97. (25) King, R.; Bonfiglio, R.; Fernandez-Metzler, C.; Miller-Stein, C.; Olah, T. J. Am. Soc. Mass Spectrom. 2000, 11, 942. (26) Bonfiglio, R.; King, R.; C.; Olah, T.; Merkle K. Rapid Commun. Mass Spectrom. 1999, 13, 1175. (27) Liu, M.; Bryant, M. S.; Chen, J.; Lee, S.; Yaremko, B.; Lipari, P., Malkowski, M.; Ferrari, E.; Nielsen, L.; Prioli, L.; Dell, J.; Sinha, D.; Syed, J.; Korfmacher, W. A.; Nomeir, A.; Lin, C.; Wang, L.; Taveras, A.; Doll, R.; Njoroge, G.; Mallams, A.; Remiszewski, S.; Catino, J.; Girijavallabhan, V.; Kirschmeier, P.; Bishop, R. Cancer Res. 1998, 58, 4947.
Table 1. HPLC Flow Programming
time (min) 0.00 0.01 0.45 0.50 0.60 0.70 1.20 1.30
composition of mobile phase (%) A B 100 100 100 10 10 0 0 100
0 0 0 90 90 100 100 0
flow rates (mL/min)
valve positionsa
4.0 8.0 8.0 6.0 4.0 1.2 1.2 4.0
B B B B B A A B
a B, HPLC flow diverted to waste; A, HPLC flow diverted to the mass spectrometer.
Figure 1. Direct monolithic column MRM chromatograms of I after 10-, 20-, 30-, and 40-µL plasma injection.
Mobile phases A and B were 4 mM ammonium acetate in wateracetonitrile (90:10) and 4 mM ammonium acetate in wateracetonitrile (10:90) containing 0.015% TFA, respectively. Equipment. HPLC-MS/MS analysis was performed using a PE Sciex (Concord, ON, Canada) model API 3000 triple quadrupole mass spectrometer equipped with a heated nebulizer interface. The HPLC system consisted of a Leap autosampler with a refrigerated sample compartment (set to 10 °C) from LEAP Technologies (Carrboro, NC), on-line degasser, LC-10AD VP pump, and LC-10A VP controller from Shimadzu (Columbia, MD). For the protein precipitation assay, a Synergi C18 column (2.0 × 30 mm, 4 µm) from Phenomenex Inc. (Torrance, CA) was used as the analytical column.8 For the monolithic column method, a Chromolith SpeedROD, RP-18e (4.6 × 50 mm) from Merck KgaA (Darmstadt, Germany) was used. The Quadra 96, model 320 (Tomtec, Hamden, CT) was used for semiautomated sample preparation for the protein precipitation method. For the matrix ionization suppression studies, a mixture containing the test compounds I and II and the ISTD was continuously infused into Peek tubing between the monolithic silica column and mass spectrometer through a tee using a Harvard Apparatus model 2400 (South Natick, MA) syringe pump as described elsewhere.4,7,8 Standard and Sample Preparation. Stock solutions of I, II, and the ISTD were prepared as 1 mg/mL solutions in methanol. Analytical standard samples were prepared by spiking known quantities of the standard solutions to blank plasma. The concentration ranges for both analytes in plasma were 1-2500 ng/mL level. Study blood samples were collected at specified time points following oral administration to a single animal. After clotting on ice, plasma was isolated by centrifugation and stored frozen (-20 °C) until analysis. For the protein precipitation procedures, a 150µL aliquot of an acetonitrile solution containing 1 ng/µL ISTD was added to 50 µL of plasma. After vortex mixing and centrifugaAnalytical Chemistry, Vol. 75, No. 8, April 15, 2003
1813
Figure 2. Peak responses of I, II, and the ISTD as a function of injection volumes.
tion, the supernatant was transferred to a new 96-well plate using the Quadra 96 system.8 Aliquots of 10 µL were injected for HPLCMS/MS analysis. For the direct injection method, 50 µL of plasma was loaded into a 96-well plate and diluted with 200 µL of deionized water containing the ISTD at a concentration of 1 ng/µL. Then 10-µL aliquots of the diluted plasma were directly injected into the monolithic column. Chromatographic Conditions. Chromatographic separation for the regular silica particle-packed column was achieved using a two-solvent gradient system as described previously.8 The instrumental configuration of the single-column direct HPLCMS/MS method was reported elsewhere.18 A 10-µL portion of the diluted plasma sample was injected by the autosampler into the monolithic silica C18 column. The switching valve was first diverted to waste during the removal of the macromolecules from plasma matrix. The valve was then switched to the mass spectrometer, and a fast gradient from 0 to 90% B was initiated to elute and separate the analytes. The separation stages were followed by the equilibration stage with the divert valve switched back to the waste and mobile phase changed from B to A. The HPLC flow programming is described in Table 1. Mass Spectrometric Conditions. The mass spectrometer was operated in positive ion mode. The heated pneumatic nebulizer probe conditions were as follows: 500 °C temperature setting, 80 psi nebulizing gas pressure, 1.0 L/min auxiliary gas flow, 0.9 L/min curtain gas flow rate. The product ion spectra of I, II, and the ISTD were reported previously.9,20 The MS/MS transition selected to monitor the I was from m/z 493 (MH+) to a product ion at m/z 447. The protonated molecules were fragmented by collision-activated dissociation with nitrogen as collision gas at a pressure of instrument setting 4. The collisionoffset voltage was set at 40 V. The amine metabolite and the ISTD were monitored using the transitions from m/z 421 to 404 and m/z 639 to 471, respectively. The experimental mass spectrometric conditions were determined using a generic state file. 1814
Analytical Chemistry, Vol. 75, No. 8, April 15, 2003
RESULTS AND DISCUSSION Development of Fast Direct Plasma Injection Method. The objective of the study was to investigate the utility of a monolithic silica column for direct HPLC-MS/MS analysis. The dual role of the monolithic column is both to remove matrix macromolecules and to provide chromatographic efficiency comparable to a conventional microparticulate silica column for small molecules.28-35 Although the potential of using the monolithic silica column for direct plasma injection was demonstrated previously, inconsistent column efficiencies following plasma analyses and poor performance in sensitivity were observed.34 In general, the monolithic column and a regular microparticulate reversed-phase column provide similar chromatographic performance.9,31,35 In contrast to the particle-packed silica column, the monolithic column generates much flatter Van Deemter plots at high flow rates due to the better mass-transfer properties of a monolithic skeleton versus distinct particles allowing for faster HPLC separations without a noticeable effect on chromatographic resolution.9,28 However, the use of higher (>2 mL/min) flow rates for further reduction of analysis time will typically result in a lower signal when electrospray ionization or heated nebulizer is used as an ionization source for the mass spectrometer.9,33 One solution is to utilize postcolumn splitting, which may result in poorer assay (28) Tanaka, N.; Kobayahi, H.; Nakanishi, K.; Minakuchi, H.; Ishizuka, N. Anal. Chem. 2001, 73, 421A. (29) Cabrera, K.; Wieland, G.; Lubda, D.; Nakanishi, K.; Soga, N.; Minakuchi, H.; Unger, K. K. Trends Anal. Chem. 1998, 17, 50. (30) Dear, G.; Plumb, R.; Mallett, D. Rapid Commun. Mass Spectrom. 2001, 15, 152. (31) Bidlingmaier, B.; Unger, K. K.; Doehren, N. v. J. Chromatogr., A 1999, 832, 11. (32) Wu, J.; Zeng, H.; Deng, Y.; Unger, S. E. Rapid Commun. Mass Spectrom. 2001, 15, 1113. (33) La¨mmerhofer, M.; Svec, F.; Frechet, J. M. J.; Lindner, W. Anal. Chem. 2000, 72, 4623. (34) Plumb, R.; Dear, G.; Mallett, D.; Ayrton, J. Rapid Commun. Mass Spectrom. 2001, 15, 986. (35) McCalley, D. V. J. Chromatogr., A 2002, 987, 17.
Figure 3. Comparison of chromatographic performance at the first (solid line) and 192th (dotted line) injections of diluted rat plasma.
sensitivity.9,33 As an alternative, we chose to use flow programming without postcolumn splitting as the solution. For the optimization of HPLC separation, the flow rate is programmed to start at high flow rate and adjusted to a low flow rate during the retention times of the analytes, as summarized in Table 1. The performance of the monolithic silica column was examined by using a mixture containing I, II, and the ISTD throughout the experiments. Figure 1 shows the reconstructed MRM chromatograms of I after 1040-µL injections of the diluted plasma. Figure 1 demonstrates that the retention time and bandwidth remain unchanged with increasing injection volumes. This phenomenon was also observed for II and the ISTD (data not shown). The linear relationship of all three compounds based on peak areas up to 40-µL injection volumes is shown in Figure 2. These data indicate a loading capacity of at least 40 µL of diluted plasma into the monolithic column. The concentration of each of these compounds in the undiluted rat plasma sample was 250 ng/mL.
The ruggedness of the monolithic column was studied by sequential 10-µL injection of 200 diluted plasma samples. The diluted plasma was equally distributed into two 96-well plates as routinely used in our laboratory for HPLC-MS/MS analysis. The response ratio of I and II versus the ISTD against injection number has overall 13.5 and 10.1% coefficients of variation (CVs), respectively. The retention times of I, II, and the ISTD were reproducible with 1.9, 1.9, and 3.5% (CV), respectively. The column back pressure increased less than 5% over the last batch of plasma samples based on the HPLC conditions described in Table 1. The representative reconstructed MRM chromatograms for these compounds from the first (solid line) and the 192th (dotted line) injection, shown in Figure 3, indicate the consistent peak shapes and retention times are achievable under the proposed chromatographic conditions. The duration of the initial chromatographic stage for sample cleaning was found to be critical for a prolonged usage of monolithic silica column. The MS/MS transitions for clozapine and other test compounds are indicated in the Figure 4. Figure 4 shows the direct monolithic column MRM chromatograms of clozapine, an antipsychotic agent, and other test compounds III-XI in rat plasma, demonstrating the capability for simultaneous determination of small molecules. The test compounds III-XI are representative new chemical entities from three different drug discovery programs with molecular weights ranging from about 400 to 600. Among these compounds, IV is the hydroxyl metabolite of III; as shown in Figure 4, baseline chromatographic separation was achieved between those two analytes. The column efficiencies for these test compounds ranged from 20 000 to 63 000 theoretical plates/m around 1-min run time. A comparison of the peak area responses of the test compounds in the spiked plasma with those from the spiked supernatant solution with the protein precipitation technique provided
Figure 4. Direct monolithic column MRM chromatograms of clozapine and test compounds III-XI.
Analytical Chemistry, Vol. 75, No. 8, April 15, 2003
1815
Table 2. Accuracy and Precision Data for Compounds I and II in Rat Plasma interday 1
interday 2
norminal concn (ng/mL)
mean daily accuracya (%)
precisiona (CV, %)
mean daily accuracya (%)
10 50 250 1000
102 97.1 102 100
6.5 4.7 3.0 3.7
95.6 102 100 103
10 50 250 1000
98.3 98.6 104 106
6.4 8.1 7.8 2.8
101 92.2 91.7 93.4
a
interday 3
precisiona (CV, %)
mean daily accuracya (%)
precisiona (CV, %)
interday accuracy mean recoveryb (%)
Compound I 7.3 5.4 3.4 5.7
92.8 90.3 93.7 99.9
7.0 2.4 6.5 2.0
96.8 96.5 98.6 101
Compound II 8.0 5.1 4.4 4.1
102 97.1 102 100
6.5 4.7 3.0 3.7
100 96.0 99.2 99.8
Intraday, n)6. b Intraday, n)18.
Figure 6. Calibration plots of (A) I and (B) its amine metabolite using direct monolithic column HPLC-MS/MS method. Figure 5. Infusion MS/MS chromatograms of (A) ISTD, (B) I, and (C) II after injection of mobile phase (dotted line) or plasma (solid line).
an indication of recovery for each drug candidate for an on-line column extraction procedure, which can be referred to as “apparent on-column recovery”. The analytical recoveries of the analytes and the ISTD added to plasma were studied with the rat plasma (diluted 1:4 with water) spiked at a 250 ng/mL concentration level. The calculated recovery values of the test compounds I and II (its amine metabolite) and the ISTD (five sample injections) were found to be 99.8 (CV ) 2.4%), 99.9 (CV ) 2.7%), and 95.1% (CV ) 5.1%) in rat plasma, respectively. These values were reproducible and acceptable for drug analysis. Individual and mean accuracy values for the test compound and its amine metabolite and corresponding CVs for all QC samples are given in Table 2. The intraday mean accuracy values for both analytes ranged from 90.3 to 106%, and the intraday precision for both analytes was less than 10% CV. Investigation of Matrix Ionization Suppression. A wellaccepted concern about assay reliability when rapid HPLC-MS/ MS methods are being developed is the increased likelihood of encountering matrix ionization suppression problems.8,23 As reported, matrix ionization suppression appears to be more likely a problem when the protein precipitation method is used for sample preparation as compared to the liquid-liquid extraction or solid-phase extraction methods.25,26 Despite this concern, the protein precipitation method has been chosen as the standard sample preparation procedure for HPLC-MS/MS assays in our 1816 Analytical Chemistry, Vol. 75, No. 8, April 15, 2003
laboratory due to its simplicity.2 The matrix ionization suppression effects when monolithic column HPLC-MS/MS techniques were used with protein precipitation extracts of rat plasma samples was previously explored by monitoring the variability of the ionization responses for the test compound, its metabolite, and the ISTD using the postcolumn infusion scheme in our laboratory.9 Previous data concluded that ionization suppression occurred at the early phase of a given chromatographic run, and the length of time required for the response to return to its preinjection signal levels was shortened with increasing flow rate.9 In this investigation, we focused on exploring the impact of matrix ionization suppression effects when employing the monolithic column for the direct HPLC-MS/MS system. The variation in the infusion MRM chromatograms between the mobile-phase injection and the plasma extract injection was assumed to be caused by ionization suppression due to plasma sample extract constituents such as salts and nonvolatile polar molecules.25 For accurate quantitative determinations, it is recommended that the retention times of all analytes should be in the chromatographic region of little or no matrix ion suppression when the HPLC-MS/MS methods are used.24 These data provide some information about the time profile of the interference eluting from the column and the ability of the monolithic column to remove endogenous plasma components that cause the changes in the observed ionization response of the analytes. Figure 5 shows the infusion MRM chromatograms of I, II, and the ISTD monitored after the divert valve was switched from waste to the mass spectrometer. The increase in the ionization efficiencies of all three compounds over the displayed
Figure 7. Plasma concentration profiles of (A) I and (B) its amine metabolite obtained by direct and indirect monolithic column HPLC-MS/MS method and traditional particle-packed silica column HPLC-MS/MS method.
time frame can be attributed to two factors: the composition of the mobile phase and the eluent flow rate. In Figure 3, the increasing response for all three compounds was observed as the combined result of higher organic content of the mobile phases and the decreasing flow rate. It can be seen that there was no significant difference in the infusion MRM chromatograms obtained following injection of mobile phase versus plasma. These data demonstrate that little or no matrix ion suppression would be seen for both analytes and the ISTD when the direct monolithic HPLC-MS/MS method was used. Pharmacokinetic Screening. The use of the proposed monolithic column HPLC-MS/MS method was further applied for the simultaneous determination of the dosed compound (I) and its amine metabolite (II) from a pharmacokinetic experiment to demonstrate the suitability of fast direct analyses for actual drug discovery samples. The dosed compound and its amine metabolite were simultaneously monitored by the proposed direct analysis method with a baseline resolution of Rs )3.6 within 1-min run time without traditional sample preparation procedures. The separation efficiencies of the dosed compound and its amine metabolite were approximately 22 000 and 39 000 theoretical plates/m, respectively. It is important to be able to simultaneously assay for metabolites because the data obtained may help explain the observed pharmacokinetic or toxicological behavior as well as suggest further chemical structure modifications for drug discovery programs.1,2 The HPLC-MS/MS chromatograms of drug-free blank rat plasma did not show an interference or endogenous response for these analytes or the ISTD (data not shown). Figure 6 shows that the calibration curves obtained from
duplicate standard samples at each concentration level were linear with a correlation coefficient r2 ) 0.999 and 0.997 for the dosed compound and its amine metabolite, respectively. Accuracy (% bias) was less than 15% at all concentrations, 1-2500 ng/mL. The concentration-time profiles of the dosed compound and the metabolite found in the rat plasma following an oral administration at 10 mg/kg dosing are shown in Figure 7. We examined the accuracy of the proposed direct injection method by comparing the analytical results obtained by the protein precipitation method using the monolithic and particle-packed silica column and the proposed direct plasma injection method. The Student t test results also indicate no significant difference of plasma concentrations at each time point of the dosed compound and its M-72 metabolite for the three aforementioned assays with 95% confidence (R ) 0.5). The areas under the curve, AUC(0f48), of the dosed compound from the proposed direct plasma injection method and the protein precipitation method using the monolithic and the particle-packed silica column were 8550 (h × ng/mL), 8720 (h × ng/mL) (1.9% difference from the direct analyses method), and 8370 (h × ng/ mL) (2.1% difference from the direct analyses method), respectively. The AUC(0f48) of the metabolite from the proposed direct plasma injection method and the protein precipitation method using the monolithic and the particle-packed silica column were 1130 (h × ng/mL) and 1160 (h × ng/mL) (2.7% difference from the direct analyses method), respectively. The above results show that the proposed direct analysis method is equivalent to the nondirect injection methods in terms of accuracy. However, the proposed approach is simpler and more efficient over the method previously reported and significantly reduces the off-line sample Analytical Chemistry, Vol. 75, No. 8, April 15, 2003
1817
preparation time to provide high-throughput biological sample analysis for pharmacokinetic evaluation. CONCLUSIONS An efficient bioanalytical method based on monolithic silica column HPLC-MS/MS for on-line purification and separation has been demonstrated for the simultaneous direct determination of a drug candidate and its amine metabolite in rat plasma within a 1-min run time. The proposed method should prove to be reproducible, sensitive, and applicable to high-throughput PK screening applications and should speed the process by reducing tedious steps such as sample preparation. The on-line extraction procedure using the silica “rod” column was designed both for the removal of unwanted plasma matrix without producing system-
1818
Analytical Chemistry, Vol. 75, No. 8, April 15, 2003
clogging and ionization suppression problems and to simultaneously provide traditional analytical chromatography for the test compound and the metabolite for more than 200 injections of the diluted rat plasma samples. ACKNOWLEDGMENT The authors thank our Pharmacology Department for planing the animal studies presented in this work.
Received for review October 10, 2002. Accepted February 14, 2003. AC020630E
