Anal. Chem. 2009, 81, 9002–9011
Development of an Information-Rich LC-MS/MS Database for the Analysis of Drugs in Postmortem Specimens Hsiu-Chuan Liu,†,‡ Ray H. Liu,| Hsiu-O Ho,*,† and Dong-Liang Lin*,‡,§,⊥ College of Pharmacy, Taipei Medical University, Taipei 110, Taiwan, Department of Forensic Toxicology, Institute of Forensic Medicine, Ministry of Justice, Taipei 106, Taiwan, Department of Medical Technology, Taipei Medical University, Taipei 110, Taiwan, Department of Medical Technology, Fooyin University, Kaohsiung Hsien 831, Taiwan, and Department of Forensic Medicine, College of Medicine, National Taiwan University, Taipei 100, Taiwan With several instrument configurations available from various manufacturers, liquid chromatography-tandem mass spectrometry (LC-MS/MS) technology is currently intensively studied for comprehensive drug screen and confirmation. An LC-MS/MS database, including 780 drug and toxic compounds, has been constructed, featuring information-rich MS/MS spectra derived from a novel fragmentation approach incorporating voltage ramping and broadened mass window for activation. The resulting spectra are rich in high- and low-mass fragment ions, highly effective for matching and proven reproducible over a 6 month test period. Coupled to effective sample preparation protocols, the database-searching process greatly improved the identification of drugs in postmortem specimens by the LC-electrospray ionization (ESI)-MS/ MS technology. This method has significantly improved the efficiency of our routine laboratory operation that was based on a two-step [fluorescence polarization immunoassay (FPIA) and gas chromatography/mass spectrometry (GC/MS)] approach in the past. With the availability of instrumentation in varied configurations from different manufacturers, liquid chromatography-tandem mass spectrometry (LC-MS/MS) is now intensely studied for its utilization in preliminary screen and confirmatory analysis of various categories of analytes, including drugs of abuse.1 The goal is to achieve the same effectiveness currently provided by the mature and robust gas chromatography/mass spectrometry (GC/ MS) technology and to go beyond this level by taking advantages of the LC’s ability in handling compounds of lower volatility and the intrinsic characteristics of the tandem MS technology. To reach this perceived potential, one must have available an ionization method and a platform (for the analysis of the resulting * To whom correspondence should be addressed. E-mail:
[email protected] (D.-L.L.);
[email protected] (H.-O.H.). Phone: +886-2-27392369, ext 700 (D.-L.L.); +886-2-27361661, ext 6126 (H.-O.H.). Fax: +886-2-27360875 (D.-L.L.); +886-2-27390671 (H.-O.H.). † College of Pharmacy, Taipei Medical University. ‡ Institute of Forensic Medicine, Ministry of Justice. § Department of Medical Technology, Taipei Medical University. | Department of Medical Technology, Fooyin University. ⊥ Department of Forensic Medicine, National Taiwan University. (1) McCurdy, H. H.; Morrison, A. M.; Holt, L. A. Forensic Sci. Rev. 2008, 20, 45–73.
9002
Analytical Chemistry, Vol. 81, No. 21, November 1, 2009
ions) that are capable of generating information-rich mass spectra for the analytes of interest. The resulting mass spectra must be reproducible (at different times and different laboratories) with the analytes at a very low concentration level. A significant number of reports have indicated that electrospray ionization (ESI) method can achieve better sensitivity,2 especially for polar and thermolabile compounds,3 although it is also more susceptible to the ion suppression phenomenon.4,5 Whether the ion trap or the triple-quadrupole configuration is a better platform highly depends on the nature of the test sample and the analytical objective. There have also been a substantial number of studies on various operation parameters, such as collision energy,6-11 collision gas pressure,7 and dwell time,9,12 striving to attain the desirable sensitivity and reproducibility of multiple reaction monitoring (MRM) spectra for all compounds of interest at the same time. One essential component contributing to the effectiveness of the GC/MS technology is the availability of library database consisting the analytes’ mass spectra that can be routinely reproduced by average laboratories around the world. Thus, there have been reports6-13 on the creation of a searchable LC-MS database generated by different ionization methods and instrumentation configurations and/or manufacturers. Lacking the universally adopted conventions enjoyed by the GC/MS methodology, these studies utilized various combinations of ionization (2) Scha¨nzle, G.; Li, S.; Mikus, G.; Hofman, U. J. Chromatogr., B 1999, 721, 55–65. (3) Maurer, H. H. J. Chromatogr., B 1998, 713, 3–25. (4) Smith, M. L.; Vorce, S. P.; Holler, J. M.; Shimomura, E.; Magluilo, J.; Jacobs, A. J.; Huestis, M. A. J. Anal. Toxicol. 2007, 31, 237–253. (5) LeBeau, M. A.; Montgomery, M. A.; Wagner, J. R.; Miller, M. L. J. Forensic Sci. 2000, 45, 1133–1141. (6) Marquet, P.; Saint-Marcoux, F.; Gamble, T. N.; Leblanc, J. C. Y. J. Chromatogr., B 2003, 789, 9–18. (7) Josephs, J. L.; Sanders, M. Rapid Commun. Mass Spectrom. 2004, 18, 743– 759. (8) Dresen, S.; Kempf, J.; Weinmann, W. Forensic Sci. Int. 2006, 161, 86–91. (9) Gergow, M.; Ojanpera¨, I.; Vuori, E. J. Chromatogr., B 2003, 795, 41–53. (10) Mueller, C. A.; Weinmann, W.; Dressen, S.; Schreiber, A.; Gergov, M. Rapid Commun. Mass Spectrom. 2004, 19, 1332–1338. (11) Baumann, C.; Cintora, M. A.; Eichler, M.; Lifante, E.; Cooke, M.; Przyborowska, A.; Halket, J. M. Rapid Commun. Mass Spectrom. 2000, 14, 349– 356. (12) Zumwalt, M.; Goodley, P. C. Application Note 5988-3619EN; Agilent Technologies: Santa Clara, CA, 2002. (13) Herrin, G. L.; McCurdy, H. H.; Wall, W. H. J. Anal. Toxicol. 2005, 29, 599–606. 10.1021/ac901599d CCC: $40.75 2009 American Chemical Society Published on Web 09/29/2009
Analytical Chemistry, Vol. 81, No. 21, November 1, 2009
9003
9004
Analytical Chemistry, Vol. 81, No. 21, November 1, 2009
Figure 1. MS/MS spectra of 24 drugs exemplifying the information-rich nature of the 780 compound database. Table 1. Database and Sample Information and Search Results over a 6 Month Retention Time (RT), Fit (F), Reverse Fit (RF), Purity (P), and Molecular Weight (M) database information
test sample information
matching scores
drugs
RT
M+1
n
RT
M+1
F
RF
P
morphine codeine 6-acetylmorphine norketamine 6-acetylcodeine heroin zolpidem 7-aminoflunitrazepam bromazepam chlorpromazine clonazepam estazolam triazolam diazepam prazepam
2.1 5.5 8.5 10.2 12.7 12.9 15.0 15.0 16.5 16.9 17.5 18.3 18.8 19.1 20.3
286.4 300.5 328.5 224.5 342.4 370.4 308.3 284.3 317.3 319.8 316.7 295.8 344.1 285.7 325.8
6 6 6 6 6 6 6 6 6 6 6 6 6 6 6
2.35 ± 0.05 5.48 ± 0.05 8.56 ± 0.17 10.35 ± 0.13 12.65 ± 0.05 12.85 ± 0.05 15.03 ± 0.10 15.11 ± 0.14 16.65 ± 0.10 16.98 ± 0.07 17.56 ± 0.12 18.15 ± 0.05 18.91 ± 0.11 19.23 ± 0.08 20.48 ± 0.07
286.48 ± 0.07 300.53 ± 0.05 328.55 ± 0.08 224.70 ± 0.12 342.51 ± 0.11 370.51 ± 0.13 308.53 ± 0.08 284.35 ± 0.18 317.20 ± 0.38 319.86 ± 0.08 316.78 ± 0.13 295.53 ± 0.12 343.96 ± 0.20 285.65 ± 0.22 325.83 ± 0.05
949.6 ± 9.8 984.6 ± 4.8 973.8 ± 13.9 992.8 ± 5.4 966.6 ± 3.8 967.6 ± 12.1 974.3 ± 10.5 861.6 ± 34.7 939.5 ± 5.4 942.3 ± 19.8 990.1 ± 1.8 976.8 ± 17.4 972.5 ± 21.5 982.6 ± 4.7 997.0 ± 1.6
944.8 ± 10.0 985.6 ± 5.0 974.3 ± 13.9 993.5 ± 5.8 966.6 ± 5.5 968.6 ± 12.2 970.5 ± 5.2 859.6 ± 35.9 942.5 ± 4.8 928.6 ± 22.0 989.1 ± 2.7 965.6 ± 30.7 969.8 ± 23.4 982.3 ± 5.0 997.0 ± 1.6
944.0 ± 9.8 984.6 ± 5.0 973.0 ± 14 992.8 ± 5.4 965.1 ± 5.5 966.5 ± 12 969.8 ± 5.1 859.0 ± 37 938.6 ± 5.5 932.6 ± 29 988.8 ± 2.8 962.6 ± 28 969.6 ± 23 981.3 ± 5.0 997.0 ± 1.6
methods, instrumentation configurations, and scan modes facilitated by software development. Since the collision energy required
to attend the best signal intensity and the best mass spectral information varies with analytes, creative approaches included Analytical Chemistry, Vol. 81, No. 21, November 1, 2009
9005
Figure 2. MS/MS spectra of 7-aminoflunitrazepam obtained during a 6 month period: (A) Aug 23, 2008; (B) Sept 26, 2008; (C) Oct 23, 2008; (D) Nov 19, 2008; (E) Jan 20, 2009; (F) Feb 12, 2009.
normalizing mass spectra resulting from the use of two,6 three,7-10 or “wide-band”11 collision energies. The objective is to generate high-intensity mass spectra for all analytes that can achieve good fit to the standard spectra stored in the database. With thorough understanding of the literature data currently available, we have carried out this study, adapting the collisioninduced dissociation (CID)-MS technology in which the fragmentation voltage was ramped from 30% to 200% of a preset voltage (the “SmartFrag feature” by Agilent Technologies) and adjustable fragmentation width, designed “to fragment at the same time both the precursor ion and any product ions resulting from low-energy fragmentation pathways, thus avoiding the need for further stages 9006
Analytical Chemistry, Vol. 81, No. 21, November 1, 2009
of MS.”14 The resulting product ion spectra were found significantly richer; thus, a library was constructed based on the analytes’ retention data and the information-rich MS/MS mass spectra of 780 standard drugs and toxic compounds. Data resulting from these analytes in a 6 month test period were found to be remarkably reproducible in mass spectral information as demonstrated by the NIST-based search algorithm. The established analytical protocol and the library database were further evaluated by limit of detection (LOD) studies of a set of representative analytes and their application to the analyses of three categories of samples commonly encountered in toxicological laboratories. The LC “fragment-rich” MS/MS library system was found highly
Figure 3. MS/MS spectra of 6-acetylmorphine derived from urine specimens with the analyte’s concentration ranging from 100 to 5 ng/mL.
effective for broad preliminary screen and confirmatory analysis of drugs and toxic compounds. The specificity and sensitivity of this method have significantly improved the efficiency of our laboratory operation that was previously based on the immunoassay and GC/MS two-step approach. MATERIALS AND METHODS Chemicals and Reagents. All solvents and reagents were HPLC grade and purchased from J. T. Baker Inc. (Phillipsburg, NJ). All standard drug solutions (1.0 mg/mL in methanol) were provided by Cerilliant Corporation (Austin, TX) or Sigma-Aldrich Corporation (St. Louis, MO). Urine and blood confirmed negative by LC-MS were used to prepare standards and controls.
Solid-phase extraction (SPE) sorbent Isolute HCX (130 mg) was obtained from International Sorbent Technology Ltd. (Mid Glamorgan, U.K.). SPE was done with a vacuum manifold processing station from Agilent Technologies (Santa Clara, CA). Toxi-Tube A from Varian (Harbor City, CA) was used for liquid-liquid extraction (LLE). Instrumentation. The LC-MS/MS system consisted of an Agilent LC/MSD Trap XCT mass spectrometer fitted with an electrospray interface and an Aglilent 1100 HPLC system (Santa Clara, CA). Chromatographic separation was achieved using an Agilent Zorbax SB-Aq (2.1 mm × 150 mm, 3.5 µm particle) analytical column operated at 40 °C. The mobile phase consisted of 0.1% formic acid (v/v) in water (A) and methanol (B) at a flow Analytical Chemistry, Vol. 81, No. 21, November 1, 2009
9007
Table 2. Limits of Detection of 24 Drugs Derived from Automatic Searching of the Information-Rich LC-MS/MS Database LOD (ng/mL) urine
blood
compound
RT (min)
MW (m/z)
MS1 ion (m/z)
MS2 ion (m/z)
SPE
LLE
SPE
LLE
morphine-3β-glua normorphine hydromorphone-3β-glu dihydromorphine morphine morphine-6β-glu noroxymorphone oxymorphone hydromorphone nalorphine norcodeine dihydrocodeine naloxone codeine noroxycodone oxycodone naltrexone hydrocodone 6-acetylmorphine 6β-naltrexol 6-acetylcodeine heroin norbuprenorphine buprenorphine
1.7 1.9 2.0 2.1 2.2 2.2 2.5 2.6 3.1 4.9 5.1 5.3 5.4 5.7 7.0 7.3 8.1 8.6 8.9 9.4 12.8 13.0 15.1 16.4
461.5 271.3 461.5 287.4 285.3 461.5 287.3 301.3 285.3 311.4 285.3 301.4 327.4 299.4 301.3 315.4 341.4 299.4 327.4 343.4 341.4 369.4 413.6 467.7
462.5 272.3 462.5 288.3 286.4 462.4 288.3 302.5 286.4 312.4 286.4 302.4 328.4 300.5 302.3 316.3 342.6 300.5 328.5 344.5 342.4 370.4 414.6 468.6
286.1 254.2 286.1 187.0 201.2 286.1 270.2 284.3 185.2 270.2 268.2 200.5 310.3 215.1 284.2 298.3 324.4 199.0 211.0 326.2 225.1 328.2 396.3 414.3
NDb 50 ND 25 20 ND 50 10 20 20 10 10 20 10 20 20 5 10 20 1 10 10 5 2.5
ND 500 ND 50 20 ND 200 5 20 10 10 5 2.5 10 50 50 10 10 5 2.5 5 5 20 2.5
ND 50 ND 25 20 ND 50 10 20 20 20 10 10 10 20 10 5 20 20 1 10 10 20 2.5
ND 1000 ND 50 20 ND 500 5 20 10 20 5 20 10 50 20 2.5 20 20 2.5 10 20 50 5
a
glu: glucuronide. b ND: not detected.
rate of 0.31 mL/min. The initial gradient composition (90% A/10% B) was held for 3 min, then decreased to 0% A in 18.5 min and held for 2.5 min, then increased to 90% A in 1 min. For recycling, the initial gradient composition was restored and allowed to equilibrate for 6 min. The electrospray source was operated at 350 °C with an ionization voltage of 3500 V in positive mode. Gas source (nitrogen) was via an Agilent oxygen analyzer. The nebulizer gas pressure and the drying gas flow rate were set at 40 psig and 10 L/min, respectively. Mass spectrometric analysis was performed in positive-ion mode, performing the precursor ion scans with the following parameters: m/z range, 100-500; ramped collision energy 0.3-2.0 V (SmartFrag); precursor ion isolation width, 4 amu; spectra acquired, one MS and one MS/MS. Sample Preparation. Standard solutions of the analytes in 1-10 µg/mL were prepared using the initial gradient solvent system, A/B 90:10 (v/v). An amount of 10 µL of each standard solution was injected for the generation of library data. For SPE operation, the manufacturer’s instructions were followed. Typically, 1 mL of urine or blood sample was adjusted to pH 6.0 with 1 mL of 0.1 M phosphate buffer. The SPE column was conditioned by the addition of 1 mL of methanol, 1 mL of deionized water, and followed by 1 mL of phosphate buffer (pH 6.0). The sample was applied to the column with flow time g2 min. The columns was then washed with 2 mL of deionized water, 2 mL of 0.01 N HCl, and 2 mL of methanol, then dried for 2 min. Analytes were eluted with 2 mL of ethyl acetate/methanol/ ammonium hydroxide (73:25:2, v/v) mixture. The resulting extract was dried under a stream of nitrogen at 50 °C. For LLE by Toxi-Tube A, to the tube (containing sodium carbonate and bicarbonate, pH 9.0, in a mixture of dichloromethane, dichloroethane, n-heptane, and ethyl acetate) was added 1 mL of urine or blood, and it was vortex-mixed for 10 min. After centrifuga9008
Analytical Chemistry, Vol. 81, No. 21, November 1, 2009
tion, the upper organic phase was aspirated and dried under nitrogen gas. The dried residue was then reconstituted with 100 µL of the initial mobile phase solvent (A/B 90:10, v/v) for HPLC injection. Hair and fingernail samples were carefully cut into small segments. Test specimens were prepared by first spiking 50 mg of cut and decontaminated hair or fingernail was spiked with 50 µL of an internal standard solution (5 µg/mL nalorphine). Samples were then sonicated in 2 mL of 0.1 N HCl at 60 °C for 1 h. The mixture was then centrifuged, and the aqueous layer was transferred to a clean screw-top tube. The residue hair samples were sonicated in 1 mL of 0.1 N HCl at 60 °C for 1 h. The mixture was then centrifuged, and the aqueous layer was extracted by Toxi-Tube A. Detection and Identification. The retention and MS/MS spectral data of 780 toxicologically relevant compounds (therapeutic and illicit drugs, their metabolites, and endogenous substances) resulting from the gradient HPLC-MS/MS procedure were initially established by injecting respective standard solutions into the Agilent LC/MSD Trap XCT mass spectrometer. The spectra that are specific to each transition are averaged over the width of the eluting compound. The retention and spectra data for each compound are then transferred to the library. Each library entry includes the compound name, molecular weight, chemical formula, chemical structure, CAS number, precursor ion, MS/MS spectrum, and retention time. The Agilent ion trap identification software adapted the industry standard NIST-based search mechanism, including scores for fit (F), reverse fit (RFit), and purity (P). Molecular weight and retention time are provided for each spectrum match. RESULTS AND DISCUSSION Sufficient fragmentation of the analyte precursor ion (through CID), thereby generating a significant number of product ions at higher ion current intensities, facilitates the characterization of
Figure 4. Display of database searching results showing the precursor ion and the analyte and the database matching MS/MS spectra: (A) morphine, (B) 6-acetylmorphine, (C) benzoylecgonine, and (D) cocaine.
analytes and lowers their limits of detection. The innovative fragmentation approach, incorporating ramping fragmentation voltage and broadening mass window for activation (fragmentation
width), appeared to have succeeded in producing highly informative CID spectra of the analytes of interest. At the lower voltage range, higher mass fragment ions are generated, whereas at the Analytical Chemistry, Vol. 81, No. 21, November 1, 2009
9009
Table 3. Summary of Compounds Identified in 12 Case Samples by Automatic Searching of the In-House Information-Rich LC-MS/MS Database case
sample
1 2 3 4
blood vitreous vitreous blood vitreous
5 6
7
8
9 10 11
12
compounds screen and confirmation by automatic searching of the in-house information-rich LC-MS/MS database
morphine, codeine, diphenhydramine, noscapine, diazepam morphine, codeine, 6-acetylmorphine, 6-acetylcodeine morphine, codeine, 6-acetylmorphine, chlorpheniramine codeine, 6-acetylmorphine, mirtazapine, zolpidem, flurazepam morphine, codeine, chlorpheniramine, mirtazapine, flurazepam, sulpiride blood morphine, codeine, chlorpheniramine, methamphetamine blood diphenhydramine, dextromethorphan urine morphine, codeine, 6-acetylmorphine, 6-acetylcodeine, diphenhydramine, dextromethorphan, 7-aminonitrazepam gastric heroin, morphine, codeine, 6-acetylmorphine, diphenhydramine, dextromethorphan blood morphine, methamphetamine, 7-aminoflunitrazepam oral fluid heroin, morphine, codeine, 6-acetylmorphine, 6-acetylcodeine, methamphetamine syringe heroin, morphine, codeine, 6-acetylmorphine, 6-acetylcodeine, noscapine blood morphine, codeine, 7-aminoflunitrazepam urine heroin, morphine, codeine, 6-acetylmorphine, 6-acetylcodeine, 7-aminoflunitrazepam syringe heroin, morphine, codeine, 6-acetylcodeine, noscapine blood morphine, codeine, atropine urine morphine, codeine, 6-acetylmorphine, 6-acetylcodeine, atropine blood morphine, codeine gastric morphine, codeine, 6-acetylmorphine, 6-acetylcodeine blood morphine, codeine, 7-aminoflunitrazepam, zolpidem, clothiapine, desalkyflurazepam urine morphine, codeine, 7-aminoflunitrazepam, 6-acetylmorphine, zolpidem, chlorpheniramine, amphetamine, methamphetamine, hydroxyethylflurazepam gastric morphine, codeine, 6-acetylmorphine, clothiapine, zolpidem, chlorpheniramine, methamphetamine blood morphine, codeine, noscapine urine morphine, 6-acetylcodeine, noscapine gastric morphine, 6-acetylcodeine, noscapine syringe heroin, morphine, codeine, 6-acetylmorphine, 6-acetylcodeine
higher voltage range, smaller fragment ions are favored. In the meantime, the increased fragmentation width appeared to have dissociated at the same time both the precursor ion and product ions resulting from low-energy fragmentation pathways.14 We have found the resulting CID spectra of 780 drug and toxic compound highly rich in product ion information and assembled them into an LC-MS/MS database. The effectiveness of the information-rich LC-MS/MS database hereby reported has been examined at three different levels. First, reproducibilities of retention time and MS/MS data have been evaluated over a 6 month period. Second, method LODs have been studied by the analyses of drug-free urine and blood samples spiked with a set of analytes ranging from 1000 to 1 ng/mL. Third, this method was applied to the analyses of external proficiency test samples and case samples, including postmortem blood and hair and nail. The MS/MS data included in the 780 compound database are exemplified by the 24 drugs shown in Figure 1. These spectra are information-rich resulting from the innovative fragmentation approach. Reproducibility of these highly informative spectra, essential to compound identification through the library-searching approach, was studied by repeatedly injecting a 15 analyte mixture, representing different categories of compounds with retention times ranging from 2.3 to 20.3 min, over a 6 month period. Standard search results are tabulated in Table 1. The six MS/ MS spectra for the compound (7-aminoflunitrazepam) with the 9010
Analytical Chemistry, Vol. 81, No. 21, November 1, 2009
confirmed by GC/MS library none none (screened positive by FPIA: opiates 90 ng/mL) none (screened positive by FPIA: opiates 138 ng/mL) mirtazapine mirtazapine none dextromethorphan dextromethorphan, levorphanol dextromethorphan none methamphetamine codeine, 6-acetylcodeine none 6-acetylmorphine codeine atropine codeine, 6-acetylmorphine, 6-acetylcodeine none none none codeine, 6-acetylmorphine, clothiapine, zolpidem, didisethylflurazepam codeine, zolpidem none codeine, 6-acetylmorphine codeine, 6-acetylcodeine, 6-acetylmorphine heroin, codeine, 6-acetylmorphine, 6-acetylcodeine
least favorable scores, as shown in Figure 2, still exhibit remarkable similarity. To evaluate the achievable LODs of the entire analytical process, the established sample preparation/LC-MS/MS procedures were applied to the analysis of a series of drug-free urine and whole blood samples spiked with a mixture of 24 drugs (Figure 1) in the following analyte concentrations: 1000, 500, 200, 100, 50, 25, 20, 10, 5.0, 2.5, 1.0 ng/mL. Replicates were extracted and analyzed in two different days to check the reproducibility. As an example, the MS/MS spectra of 6-acetylmorphine derived from the LLE protocol of urine specimen, with the analyte’s concentration ranging from 100 to 5 ng/mL, are shown in Figure 3. The LOD data shown in Table 2 represent the lowest concentrations of these compounds that were successfully identified by the automated search process. Clearly, recovery of the adopted extraction procedure plays an important role in reaching a lower LOD for a specific analyte. In summary, the SPE method was found more effective, with the LODs for these analytes ranging from 1 to 50 ng/mL. Lower LODs can be achieved through further improvement in the sample preparation step. Finally, the established method was applied to the analyses of external proficiency test samples provided by the College of American Pathology (CAP), postmortem specimens, and hair and nail samples. CAP samples containing benzoylecgonine, cocaine, and cocaethylene in one and morphine, dihydrocodeine, codeine, and 6-acetylmorphine in the other were successfully identified.
Table 4. Summary of Compounds Identified in 12 Human Hair and Fingernail Samples by Automatic Searching of the In-House Information-Rich LC-MS/MS Database sample
hair
1
methamphetamine, morphine, codeine, 6-acetylmorphine, 6-acetylcodeine, procaine, diphenhydramine, dextromethorphan, desmethyldoxepin methamphetamine, morphine, codeine, 6-acetylmorphine, 6-acetylcodeine, diphenhydramine, chlorpheniramine, tramadol, desmethyltramadol, dextromethorphan morphine, codeine, 6-acetylmorphine, 6-acetylcodeine, procaine, trazodone, carbinoxamine, diphenhydramine morphine, codeine, 6-acetylmorphine, 6-acetylcodeine, diphenhydramine, chlorpheniramine, dextromethorphan methamphetamine, morphine, codeine, 6-acetylmorphine, ketamine, carbinoxamine, diphenidol methamphetamine, morphine, codeine, 6-acetylmorphine, 6-acetylcodeine, diphenhydramine, chlorpheniramine, propanolol, butorphanol, desmethyltramadol, tramadol, dextromethorphan, citalopram methamphetamine, morphine, codeine, 6-acetylmorphine, 6-acetylcodeine, chlorpheniramine, dextromethorphan methamphetamine, morphine, codeine, 6-acetylmorphine, 6-acetylcodeine, chlorpheniramine morphine, codeine, 6-acetylmorphine, 6-acetylcodeine, tramadol methamphetamine, morphine, codeine, 6-acetylmorphine, 6-acetylcodeine, methadone, diphenhydramine, chlorpheniramine, trazodone, desmethyltramadol, tramadol, dextromethorphan, carbinoxamine, metoclopramide methamphetamine, morphine, codeine, 6-acetylmorphine, 6-acetylcodeine, chlorpheniramine, dextromethorphan, carbinoxamine morphine, 6-acetylmorphine, 6-acetylcodeine, procaine
2 3 4 5 6
7 8 9 10
11 12
The precursor ion and the MS/MS data derived from the sample containing morphine, dihydrocodeine, codeine, and 6-acetylmorphine, as shown in Figure 4, were found to match remarkably well with the data stored in the library for the respective compounds. Summarized in Table 3 is the comparison of the drugs detected in 12 postmortem case samples using the LC-MS/MS and the standard GC/MS method routinely used in our laboratory. A total of 122 compounds were found by the ESI-MS/MS/library search method compared to only 31 by the routine GC/MS procedure. Due to levorphanol not being included in the LC-MS/MS database library, the levorphanol cannot be detected in case 6. Data resulting from the application of the established LC-MS/ MS library search protocol to the analysis of hair and fingernail samples, collected from 12 self-reported heroin abusers, are shown in Table 4. Method sensitivity of this approach allows for the comparison of drugs detected in fingernail clippings and hair samples collected at the same time. 6-Acetylmorphine and acetylcodeine were found in almost all hair samples. It is generally accepted that detecting 6-acetylmorphine in a biological sample (14) Sauvage, F.-L.; Saint-Marcoux, F.; Duretz, B.; Deporte, D.; Lachatre, G.; Marquet, P. Clin. Chem. 2006, 52, 1735–1742. (15) Cone, E. J.; Welch, P.; Mitchell, J. M.; Paul, B. D. J. Anal. Toxicol. 1991, 15, 1–7. (16) Girod, C.; Staub, C. J. Anal. Toxicol. 2001, 25, 106–111. (17) Brenneisen, R.; Hasler, F.; Wursch, D. J. Anal. Toxicol. 2002, 25, 561– 566. (18) Phillips, S. G.; Allen, K. R. J. Anal. Toxicol. 2006, 30, 370–374.
fingernail morphine, codeine, 6-acetylmorphine, 6-acetylcodeine methamphetamine, morphine, codeine, 6-acetylcodeine, tramadol, econazole, dextromethorphan morphine, codeine, 6-acetylmorphine, chlorpheniramine, carbinoxamine morphine, codeine, 6-acetylmorphine, tramadol, chlorpheniramine, dextromethorphan methamphetamine, morphine, diphenidol morphine, tramadol, dextromethorphan
methamphetamine, morphine, codeine, 6-acetylmorphine, 6-acetylcodeine, heroin, dextromethorphan methamphetamine, morphine, tramadol morphine, codeine, 6-acetylmorphine, 6-acetylcodeine morphine, codeine, 6-acetylmorphine, 6-acetylcodeine, dextromethorphan morphine, 6-acetylmorphine morphine, codeine
is an indication of heroin exposure.15,16 Acetylcodeine is a manufacturing impurity found in heroin and has been suggested as a marker for the use of illicit heroin.17,18 The ability of the LC-MS/MS library search approach, capable of detecting these compounds at very low levels in hair, facilitates the differentiation of heroin consumption from codeine, morphine, or opium preparations. CONCLUSIONS To the best of our knowledge, establishment and utilization of the MS/MS library database based on the collision energy ramping approach have not yet been reported. The LC-MS/MS library search approach based on this library database has been proven highly effective for broad preliminary screen and confirmatory analysis of drugs and toxic compounds. The specificity and sensitivity of this method have significantly improved the efficiency of our laboratory operation that was previously based on the immunoassay and GC/MS two-step approach. ACKNOWLEDGMENT This study was partially supported by Grants provided by (Taiwanese) National Science Council: NSC 97-1301-04-0301; NSC 98-1301-05-06-03; NSC 96-2113-M-242-002-MY2. Received for review July 18, 2009. Accepted September 8, 2009. AC901599D
Analytical Chemistry, Vol. 81, No. 21, November 1, 2009
9011