Anal. Chem. 2001, 73, 5450-5456
LC/MS/MS Determination of Omapatrilat, a Sulfhydryl-Containing Vasopeptidase Inhibitor, and Its Sulfhydryl- and Thioether-Containing Metabolites in Human Plasma Mohammed Jemal,* Sanaullah Khan, Deborah S. Teitz, Jacqueline A. McCafferty, and Dara J. Hawthorne
Bioanalytical Research, Clinical Discovery, Bristol-Myers Squibb Pharmaceutical Research Institute, P.O. Box 191, New Brunswick, New Jersey 08903-0191
Omapatrilat, the most clinically advanced member of a new class of cardiovascular agents, vasopeptidase inhibitors, is under development at Bristol-Myers Squibb Pharmaceutical Research Institute for the treatment of hypertension and heart failure. An electrospray LC/MS/ MS method has been developed and validated for the simultaneous determination of omapatrilat and its four metabolites in human plasma. Since omapatrilat and two of the metabolites are sulfhydryl-containing compounds, methyl acrylate was used to stabilize these compounds in human blood and plasma samples. Methyl acrylate reacted instantly with the sulfhydryl group to form a derivative that was stable in blood and plasma. Extraction of the analytes from plasma samples was achieved by semiautomated liquid-liquid extraction, where a robotic liquid handler performed the liquid-transferring steps. The mass spectrometer was operated in the negative ion selected-reaction-monitoring mode. The calibration curve ranges were 0.5-250 ng/mL for omapatrilat and one metabolite and 2.0-250 ng/mL for the other three metabolites. Omapatrilat (I, Figure 1) is the most clinically advanced member of a novel class of cardiovascular compounds called vasopeptidase inhibitors (VPIs), single molecules that simultaneously inhibit both neutral endopeptidase (NEP) and angiotensinconverting enzyme (ACE).1-5 Vasopeptidase inhibition brings about the desirable cardiovascular effects by decreasing the endogenous levels of angiotensin II while increasing the endogenous levels of atrial natriuretic peptide (ANP) and bradykinin (BK). Omapatrilat is currently in clinical development for the treatment of patients with hypertension and heart failure. * Corresponding author: (telephone) 732-519-1582; (fax) 732-519-1557; (e-mail)
[email protected]. (1) Fink, C. A. Expert Opin. Ther. Pat. 1996, 6, 1147-1164. (2) Robl, J. A.; Tripoddo, N. C.; Petrillo, E. W. In Antihypertensive Drugs; Van Zwieten, P. A., Greenlee, W. J., Eds.; Harwood Academic Publishers: Amsterdam, 1997; pp 113-212. (3) Robl, J. A.; Ryono, D. E. Expert Opin. Ther. Pat. 1999, 9, 1665-1677. (4) Burnett, J. C, Jr. J. Hypertens 1999, 17 (Suppl. 1), S37-S43. (5) Graul, A.; Leeson, P. Drugs Future 1999, 24, 269-277.
5450 Analytical Chemistry, Vol. 73, No. 22, November 15, 2001
Figure 1. Chemical structures of omapatrilat (I), its four metabolites (II-V), methyl acrylate (MA), VI (omapatrilat-MA), VII (II-MA), VIII (IIIMA), and the internal standards.
The in vivo metabolic pathways for omapatrilat include Smethylation, sulfoxidation, hydrolysis of the exocyclic amide bond, and disulfide formation.6 The chemical structures of four circulating metabolites (II-V) in human plasma are shown in Figure 1. Omapatrilat and two of the metabolites (II, III) are thiols (sulfhydryl-containing compounds) whereas the other two metabolites (IV, V) are S-methyl thioethers. This paper reports the development and validation of a method to simultaneously quantitate omapatrilat and its four metabolites in human plasma using electrospray LC/MS/MS. The method is based on the use of methyl acrylate (MA) for the stabilization of omapatrilat and the two sulfhydryl-containing metabolites in human blood and plasma. (6) Iyer, R. A.; Mitroka, J.; Malhotra, B. K.; Bonacorsi, S.; Waller, S. C.; Rinehart, J. K.; Roongta, V. A.; Kripalani, K. Drug Metab. Dispos. 2001, 29, 60-69. 10.1021/ac010532d CCC: $20.00
© 2001 American Chemical Society Published on Web 10/12/2001
EXPERIMENTAL SECTION Chemicals. Omapatrilat (I), its four metabolites (II-V), and the corresponding 2H5-labeled internal standards (Figure 1) were obtained from Bristol-Myers Squibb Pharmaceutical Research Institute. Acetic acid, ammonium acetate, dipotassium hydrogen phosphate, formic acid, and MA were obtained from Aldrich (Milwaukee, WI). High-purity (18.2 MΩ) water was obtained by passing house-distilled deionized water through the Milli-Q system (Millipore, Bedford, MA). Methyl tert-butyl ether (MTBE), ethyl acetate, HPLC grade acetonitrile, and methanol were obtained from EM Science (Gibbstown, NJ). Equipment. HPLC separation was performed on a 50 × 2 mm i.d. column packed with 5-µm Aquasil C-18 (Keystone, Bellefonte, PA). A Hewlett-Packard (Palo Alto, CA) HP 1100 column heater was used to control the column temperature. A Beckman (Columbia, MD) System Gold pump (model 128) was used for solvent delivery. A Gilson (Middleton, WI) XL233 autoinjector was used for programmed randomized sample injection from a 96well plate maintained at 4 °C. A Finnigan TSQ-7000 triple quadrupole mass spectrometer (Finnigan MAT, San Jose, CA), equipped with an electrospray interface and an octapole ion guide, was used for detection. A Packard (Meriden, CT) MultiPROBE 204DT robotic liquid-handling system was used for liquidtransfer operations. A Savant (Holbrook, NY) Speed-Vac was used for drying samples by evaporation under vacuum. Vacutainer tubes (septum-sealed tubes under vacuum) containing K3EDTA were obtained from Becton Dickinson (Franklin Lakes, NJ) for collecting blood samples. Polypropylene screw-capped sample tubes (16 × 56 mm) were obtained from Elkay Products (Shrewsbury, MA). Square-well 96-well collection plates and the sealing mats to cover the plates were obtained from Varian (Harbor City, CA). Preparation of Stock Solutions. Individual stock solutions of the sulfhydryl-containing compounds I-III were prepared by dissolving the appropriate amounts in 12.5 mL of 0.01 M dipotassium hydrogen phosphate solution that was prespiked with MA. Excess MA (∼18 times the number of moles of I, II, or III) was used to achieve complete conversion of I-III to their respective MA derivativessVI-VIII (Figure 1). The final volume of each solution was then made up to 25 mL with acetonitrile. Individual stock solutions of IV and V were prepared in 50% aqueous acetonitrile. Individual stock solutions of the five internal standards were prepared in the same manner as the respective analytes. All the stock solutions were stable at 4 °C for at least 10 months. Caution: MA should be handled cautiously. All operations involving MA and samples containing MA should be carried out in a chemical hood. Butyl gloves should be worn when handling MA. Calibration Standards and Quality Control (QC) Samples. Calibration standards were prepared by spiking the appropriate amounts of IV-VIII into human plasma. The calibration curve ranges were 0.5-250 ng/mL for VI and VIII, and 2.0-250 ng/ mL for IV, V, and VII. In the calibration curve, there were 10 different concentrations for each analyte, with each concentration in duplicate. QC samples were also prepared in human plasma. Separate analyte stock solutions were used for the preparation of calibration standards and QC samples. The low-QC sample concentrations were 6.0 ng/mL for IV-VII and 1.5 ng/mL for VI
and VIII. The mid, high, and dilution QC sample concentrations were the same for all analytess100, 200, and 2000 ng/mL, respectively. The QC samples were stored in 16 × 56 mm Elkay tubes at -70 °C until analysis. Sample Processing. All the liquid-transfer steps were performed using the robotic liquid-handling system. A 0.5-mL aliquot of each standard or QC sample was transferred to a 13 × 100 mm test tube and spiked with 50 µL of a 2H5-labeled internal standard stock solution. The resulting concentration of each internal standard was 100 ng/mL of plasma. After adding 1.0 mL of 0.1 M HCl solution and 2.0 mL of ethyl acetate, the samples were shaken for 5 min. The organic layers were then transferred to 13 × 100 mm tubes and evaporated to dryness and the residues reconstituted in 50 µL of 1.0 mM formic acid solution in 1:2 acetonitrile-water. The reconstituted samples were transferred to a 96-well collection plate for injection (15 µL) into the LC/MS/ MS system. LC/MS/MS Analysis. HPLC separation was achieved isocratically with the mobile phase containing 60% A and 40% B, where A was 1.0 mM formic acid in 5:95 acetonitrile-water and B was 1.0 mM formic acid in 95:5 acetonitrile-water. The flow rate was 0.4 mL/min, and the column temperature was maintained at 40 °C. The run time was 2.5 min. The LC effluent was all directed to the mass spectrometer. The electrospray mass spectrometer was used in the negative ionization mode. The selected reaction monitoring (SRM), operated at a unit mass resolution, involved the [M - H]- of each compound as the precursor ion. The transitions monitored were m/z 195 to 147 for IV, m/z 421 to 246 for V, m/z 493 to 407 for VI, m/z 509 to 423 for VII, and m/z 267 to 181 for VIII. The SRM transitions monitored for the corresponding internal standards were m/z 200 to 152, m/z 426 to 251, m/z 498 to 412, m/z 514 to 428, and m/z 272 to 186. The collision offset energy (Coff) values used were 15, 20, 20, 25, and 15 eV for IV-VIII, respectively. The internal standards were subjected to the same Coff as the corresponding analytes. The argon collision gas was set at 2.5 mTorr. The electrospray voltage was 4.5 kV, and the capillary temperature was maintained at 250 °C. High-purity (99.999%) nitrogen was used as both the nebulizing and the auxiliary gas. A 1/x-weighted quadratic model was used for the regression of peak area ratio of analyte/internal standard versus analyte concentration. RESULTS AND DISCUSSION Instability of Thiols in Biological Matrixes and Their Stabilization via Derivatization. The analysis of thiols in biological matrixes usually involves derivatization of the sulfhydryl groups to prevent their oxidation to disulfides. MA has recently been shown to be an excellent alkylating reagent for the fast in situ protection of thiols in biological matrixes.7-9 The alkylation reaction leads to the formation of a permanent covalent bond and the MA derivatives have desirable LC/MS/MS characteristics. The structures of MA, the MA derivatives of I and the two sulfhydryl-containing metabolites II and III (VI-VIII, respectively) are shown in Figure 1. The reaction of MA with a thiol does not bring about the formation of a new asymmetric center (7) Matsuura, K.; Takashina, H. J. Chromatogr. 1993, 616, 229-234. (8) Jemal, M.; Hawthorne, D. J. Rapid Commun. Mass Spectrom. 1994, 8, 854857. (9) Jemal, M.; Hawthorne, D. J. J. Chromatogr., B 1997, 693, 109-116.
Analytical Chemistry, Vol. 73, No. 22, November 15, 2001
5451
Table 1. Stability of Omapatrilat (I) in Human Blood and Plasma at Room Temperature and on Icea % deviation from zero time time point (min) 10 30 60 120 a
blood room temp -62 -89 -96 -97
on ice -21 -43 -69 -94
responseb
plasma room temp on ice -75 -95 -98 -98
-35 -53 -76 -96
Spiked amount of omapatrilat: 250 ng/mL of blood.
Table 2. Stability of Omapatrilat-MA (VI)a in Human Blood at Room Temperature and on Ice % deviation from zero time sample 1
Table 3. Reaction between Omapatrilat (I) and Methyl Acrylate (MA): Effect of Reaction Time and Amount of MAa
sample 2
time point
room temp
on ice
room temp
on ice
30 min 1h 2h 4h 6h
-4.8 -24 -33 -54 -61
-0.9 -1.7 -3.5 -14 -20
-15 -25 -34 -53 -61
-6.4 -3.2 -10 -7.8 -2.3
a Spiked amount of VI: 50.65 ng/mL of blood in sample 1; 506.5 ng/mL of blood in sample 2.
in the derivative produced. This is in contrast to N-ethylmaleimide (NEM), another alkylating agent used to stabilize thiols in biological matrixes,8 which always produces a derivative with a new asymmetric center. Thus, the reaction of NEM with a thiol that contains a chiral center will produce two diastereomeric derivatives. If the diastereomers are chromatographically resolved, there will be two peaks arising from the single thiol compound, which is undesirable. The degree of instability of omapatrilat in human blood and plasma was investigated at room temperature and on ice. For the determination of the blood stability profile, blood samples spiked with omapatrilat were kept at room temperature or on ice for 0, 10, 30, 60, and 120 min and then reacted with MA at the specified times to form the stable MA derivative (VI). The blood samples were immediately centrifuged at 4 °C to obtain plasma. After the internal standard was added, the plasma samples were extracted and analyzed for VI. The results in Table 1 show that omapatrilat in blood is very unstable. A separate study showed that omapatrilat is equally unstable in human plasma (Table 1). The results demonstrate the need for the stabilization of omapatrilat in both blood and plasma. The stability of VI in human blood and plasma was investigated on ice and at room temperature. For the determination of the blood stability profile, blood samples spiked with VI were kept for 0, 0.5, 1, 2, 4, and 6 h, and then the blood samples were centrifuged at 4 °C to obtain plasma. The plasma samples were then spiked with internal standard and analyzed. As shown in Table 2, VI in blood on ice is stable for at least 2 h, which is in marked contrast to the high degree of instability exhibited by omapatrilat in blood on ice. However, VI in blood at room temperature exhibits a slight degree of instability after 30 min. Thus, the reaction between MA and omapatrilat in blood must 5452 Analytical Chemistry, Vol. 73, No. 22, November 15, 2001
sample 1
sample 2
sample 3
MA amount
0.5 µL 2.5 µL 5.0 µL 10 µL 20 µL
0.0127 0.0291 0.0326 0.0334 0.0279
n/ac n/a n/a n/a n/a
0.412 1.12 1.33 1.33 1.09
reaction time
10 min 30 min 1.0 h 2.0 h 4.0 h 6.0 h
0.0305 0.0303 0.0287 0.0256 0.0238 0.0221
0.789 0.733 0.593 0.600 0.566 0.553
1.55 1.31 1.07 0.998 1.00 0.927
a Spiked amount of I: 10.27 ng/mL of blood in sample 1; 256.8 ng/ mL of blood in sample 2; 513.5 ng/mL of blood in sample 3. b Response: analyte/internal standard area ratio. c n/a, not available.
be conducted on ice in order to avoid any possibility of degradation of VI formed. In a separate experiment, it was shown that VI in human plasma is stable for at least 6 h at room temperature and for at least 24 h at 4 °C. In blood, the reaction of omapatrilat with MA, on ice, was investigated by varying the reaction time and the amount of MA used. The amount of the reagent was varied from 0.5 to 20 µL of MA per milliliter of blood while keeping the reaction time constant (10 min). The reaction time was varied from 10 min to 6 h while keeping the MA amount constant (10 µL of MA/mL of blood). After the specified reaction time, the blood samples were centrifuged at 4 °C to obtain plasma. After the internal standard was added, the plasma samples were extracted and analyzed for VI. The results in Table 3 show that the optimum reaction time is 10 min and the optimum reagent amount is 10 µL of MA/mL of blood. In a separate experiment, it was determined that there was no difference between 10- and 0-min reaction times. Thus, the reaction is considered to be instantaneous. It should be noted that, in practice, a 0-min reaction time in blood is unattainable since some time is needed to mix omapatrilat and the MA reagent in blood and then centrifuge the blood sample to obtain plasma. Reproducibility of the reaction of omapatrilat in blood with MA (10 µL of MA/mL of blood) was studied extensively over a period of 12 months using two types of MA: MA (70 µL) freshly aliquotted (from the vendor-supplied reagent bottle kept at room temperature) into a K3EDTA Vacutainer tube or MA (70 µL) aliquotted into a K3EDTA Vacutainer tube and stored at -20 °C until use. For each of the two types of MA, blood (7.0 mL) was drawn into each of five MA-spiked Vacutainer tubes and mixed with the MA and K3EDTA. For each MA type, the contents of the five Vacutainer tubes were mixed to obtain a pool of 35 mL of blood containing MA and K3EDTA. The specified amount of omapatrilat was then added to each of 2-mL aliquots from the 35mL pooled blood on ice (five aliquots for each of two concentration levels of omapatrilat and each type of MA). After a 10-min reaction time, the blood samples were centrifuged at 4 °C to obtain plasma samples. After the internal standard was added, the plasma samples were extracted and analyzed for VI. Typical results are
Table 4. Reproducibility of the Reaction between Omapatrilat (I) and Methyl Acrylate (MA) in Human Blood sample type low omapatrilat concentration (10 ng/mL blood), MA fresh from bottle; day 1 experiment low omapatrilat concentration (10 ng/mL blood), MA at -20 °C for three months; day 1 experiment high omapatrilat concentration (150 ng/mL blood), MA fresh from bottle; day 1 experiment high omapatrilat concentration (150 ng/mL blood), MA at -20 °C for three months; day 1 experiment low omapatrilat concentration (10 ng/mL blood), MA fresh from bottle; day 2 experiment low omapatrilat concentration (10 ng/mL blood), MA at -20 °C for six months; day 2 experiment high omapatrilat concentration (150 ng/mL blood), MA fresh from bottle; day 2 experiment high omapatrilat concentration (150 ng/mL blood), MA at -20 °C for six months; day 2 experiment a
mean responsea
RSD (%)
0.0341
3.2
0.0351
2.8
0.493
3.5
0.501
3.3
0.0312
4.8
0.0298
3.3
0.569
2.5
0.542
5.1
Response: analyte/internal standard area ratio.
shown in Table 4. The reaction of omapatrilat with MA was very reproducible with both types of MA. To establish that the reaction of ormapatrilat with MA in blood was quantitative, samples obtained by spiking VI into blood were compared with samples obtained by spiking equivalent concentrations of omapatrilat into blood that contained MA (10 µL/mL of blood) for the in situ derivatization of omapatrilat in blood. After 10 min following the spiking, the two types of blood samples were centrifuged at 4 °C to obtain plasma. After the internal standard was added, the plasma samples were extracted and analyzed for VI. At each concentration level, there was no difference between the responses obtained from samples spiked with MA and omapatrilat and samples spiked with VI. Thus, the reaction of omapatrilat with MA in blood is quantitative throughout the curve range. The partitioning of VI between the plasma and red blood cell (RBC) components of blood was investigated at different concentrations by spiking VI into blood kept at 4 °C. The blood samples were then centrifuged at 4 °C to obtain plasma. After the internal standard was added, the plasma samples were extracted and analyzed for VI. The plasma concentrations were compared against the blood nominal (spiked) concentrations. The mean value for the ratio of plasma concentration to the blood concentration was 1.6. Assuming a hematocrit value of 0.45, the RBC binding of VI was calculated to be 12%. Liquid-Liquid Extraction (LLE) of Plasma Samples. Early on, plasma samples were extracted with MTBE after acidification with 0.1 N HCl. However, the extraction recovery of VII was very low. Thus, the lower limit of quantitation (LLQ) achieved for this compound was only 10 ng/mL, considerably higher than the LLQs achieved for the other analytes. Subsequently, MTBE was replaced by the more polar ethyl acetate, which significantly improved the extraction recovery of VII. Another potential drawback of the use
Figure 2. Full-scan Q1 mass spectra for IV-VIII.
of MTBE is the possibility of converting VI to VII due to traces of organic peroxide that may be present in MTBE. However, our investigation showed that no such conversion occurred. The semiautomated LLE procedure used for the method has been described previously.10,11 The automation significantly reduced both the analyst’s hands-on time and the overall time of the sample preparation. Mass Spectrometry. The negative ion electrospray full-scan mass spectra of the five analytes (IV-VIII), shown in Figure 2, are dominated by [M - H]- ions. The MS/MS product ion spectra of the [M - H]- ions are shown in Figure 3, with the proposed MS/MS fragmentation pathways depicted in Figure 4. In the proposed fragmentation schemes, the logical fragments are identified; however, it should be noted that there are other possible pathways to obtain the same fragments. For instance, the fragments at m/z 373, 389, and 147 for the MA derivatives VI-VIII, respectively, could be primary fragments derived directly, in a concerted fashion, from the [M - H]- ions of the compounds, or they could be secondary fragments produced from the sulfhydrylcontaining primary fragment ions. To determine the definitive fragmentation pathways, more investigation would be needed. The relative ease of the S-MA bond cleavage, the predominant fragmentation pathway for the MA derivatives VI- VIII, was significantly different for the three analytes. The S-MA bond of VII was stronger than that of VI, which in turn was stronger than that of VIII. The strength of the S-MA bond in VII was (10) Jemal, M.; Teitz, D.; Ouyang, Z.; Khan, S. J. Chromatogr., B 1999, 732, 501-508. (11) Teitz, D.; Khan, S.; Powell. M.; Jemal, M. J. Biochem. Biophys. Methods 2000, 45, 193-204.
Analytical Chemistry, Vol. 73, No. 22, November 15, 2001
5453
Figure 3. MS/MS product ion spectra of the [M - H]- ions from IV-VIII.
manifested in the relatively high abundance of the precursor ion even at high collision energy. An increase in the collision energy entailed an extensive fragmentation of VII without an increase in the abundance of the fragment resulting from the loss of MA. It is interesting to note that the analytes with the bicyclic ring moiety (V-VII) all underwent identical rearrangement to produce the thiolactone ring. Chromatography. The 1 mM formic acid-acetonitrile-water mobile phase, without salts such as ammonium acetate or ammonium formate, was found to give excellent negative ion electrospray response. This is in line with our experience with the negative ion electrospray behavior of other carboxylic acid containing compounds.12 With the Finnigan TSQ 7000 software, it was possible to use the optimal collision energy for each SRM transition without chromatographically separating the analytes. With the run time of 2.5 min, it was possible to achieve the required number of data points (>10) to adequately define the chromatographic peaks of the five analytes and five internal standards. Lower Limit of Quantitation. The LLQ for the five analytes in plasma samples was established at 2.00 ng/mL for IV, V, and VII and 0.500 ng/mL for VI and VIII, the lowest concentrations of analytes in the calibration curve. Six different lots of control plasma samples were spiked at the LLQ level of each analyte. The LLQ samples were processed and analyzed with a standard curve and QC samples containing all five analytes, and their predicted concentrations were determined. The deviations of the mean predicted concentrations from the nominal values were within (12) Jemal, M.; Ouyang, Z.; Teitz, D. Rapid Commun. Mass Spectrom 1998, 12, 429-434.
5454 Analytical Chemistry, Vol. 73, No. 22, November 15, 2001
Figure 4. Proposed MS/MS fragmentation pathways for IV-VIII.
(11% (Table 5). Typical SRM chromatograms at the LLQ levels are shown in Figure 5. Specificity. Six different lots of control human plasma were analyzed with and without the internal standards in order to determine whether any endogenous plasma constituents interfered with the analytes or the internal standards. The degree of interference was assessed by inspection of the SRM chromatograms. No significant interfering peaks from the plasma were found at the retention times and in the ion channels of either the analytes or the internal standards.
Table 5. Accuracy and Precision analyte
LLQ (ng/mL)
nominal concn mean concn accuracy (%) precision (%) low QC (ng/mL) nominal concn mean concn accuracy (%) precision (%) mid QC (ng/mL) nominal concn mean concn accuracy (%) precision (%) high QC (ng/mL) nominal concn mean concn accuracy (%) precision (%) dil QC (ng/mL) nominal concn mean concn accuracy (%) precision (%)
VI
VII
VIII
IV
V
0.500 0.460 -7.5 15.0 1.50 1.40 -6.6 6.8 100 94.0 -6.0 2.7 200 186 -6.9 2.1 2000 1940 -2.9 3.9
2.00 1.87 -6.7 14.0 6.00 6.48 8.0 2.5 100 110 10.0 2.6 200 219 9.7 4.6 2000 2240 11.9 1.3
0.500 0.48 -4.7 9.4 1.50 1.58 5.3 6.3 100 104 4.5 2.7 200 203 1.4 1.7 2000 2110 5.6 2.9
2.00 2.15 7.3 11.7 6.00 6.11 1.8 12.2 100 104 4.3 2.1 200 209 4.4 2.5 2000 2140 6.9 3.3
2.00 1.80 -10.2 6.5 6.00 6.12 2.0 9.1 100 107 7.2 1.6 200 207 3.5 1.6 2000 2170 8.6 0.9
Accuracy and Precision. The accuracy and precision of the method was determined by analyzing QC samples at concentrations within the lower, the second, and the upper quartiles of the standard curve (low, medium, and high). A fourth QC sample, with a concentration higher than the upper limit of the standard curve range, was also analyzed. This QC sample was diluted 1:9 with control plasma before analysis. Five replicate samples at each concentration were analyzed in three separate runs. The results of the three-day validation, summarized in Table 5, show that both the assay accuracy (deviation from nominal concentration) and precision (RSD) were within 15%. Stability. Benchtop, freezer storage, and freeze/thaw stability of the five analytes (IV-VIII) in human plasma was evaluated using low-, medium-, and high-QC samples. All the analytes were stable for at least 6 h at room temperature, for at least 24 h at 4 °C, for at least six months at -70 °C, and after three freezethaw cycles. Stability of the five analytes in reconstituted samples that were kept at room temperature was also assessed using QC samples at the three concentration levels. The processed samples were stable for at least 48 h at room temperature. Collection of Postdose Blood Samples from Human Subjects. During clinical studies, blood samples from individuals dosed with omapatrilat were drawn into Vacutainer tubes containing K3EDTA and MA. For this, the commercially available K3EDTA Vacutainer tubes were fortified in our laboratory with 10 µL of MA for each milliliter of blood to be collected. The addition of the specified amount of MA was achieved by piercing through the Vacutainer tube septum using a syringe with a noncoring stainless needle. The fortified Vacutainer tubes, stored at -20 °C until the day of use, maintained their vacuum for at least one year. On each day of blood sample collection, the tubes were brought out of the freezer and kept on crushed ice. Just before drawing the blood sample, each Vacutainer tube was gently rotated on its side in order to wet the wall of the tube with the reagents. The blood was drawn directly into the sealed Vacutainer tube and then thoroughly mixed with the K3EDTA and MA by gently inverting the tube several times. The tube was then placed on
Figure 5. SRM chromatograms of IV-VIII , obtained from a plasma spiked with the analytes at the LLQ levels.
crushed ice for 10 min and centrifuged for 15 min at 1000g and 4 °C to separate the plasma. The plasma was then transferred to a polypropylene tube in a fume hood. The plasma samples were stored at -70 °C until analysis. MA causes some hemolysis of the blood collected in fortified Vacutainer tubes. Therefore, the effect of such a hemolysis on the recovery of the analytes from blood was investigated. Accordingly, samples obtained by spiking the analytes (IV-VIII) into regular blood were compared with samples obtained by spiking the analytes (IV-VIII) into blood that contained MA (10 µL/mL of blood). The blood samples were centrifuged after 10 min to obtain plasma. After adding the internal standards, the plasma samples were extracted and analyzed. The results showed that there was no difference in the measured concentrations obtained from the two sets of plasma samples. CONCLUSIONS The LC/MS/MS method presented here, based on the use of MA for the stabilization of sulfhydryl-containing analytes in blood, has successfully been used for analysis of thousands of samples from clinical studies of omapatrilat. A procedure has Analytical Chemistry, Vol. 73, No. 22, November 15, 2001
5455
been developed and implemented for the safe collection of blood samples into Vacutainer tubes fortified with the specified volumes of MA for protecting sulfhydryl-containing compounds by forming stable MA derivatives. The SRM transitions used were unique for each analyte, and hence, there was no need for chromatographic separation of the analytes. The short run time of 2.5 min, combined with the automated LLE sample preparation, allowed high-
5456
Analytical Chemistry, Vol. 73, No. 22, November 15, 2001
throughput quantitation of omapatrilat and its four metabolites, using a separate stable isotope-labeled internal standard for each analyte. Received for review May 8, 2001. Accepted September 5, 2001. AC010532D
