Application of Direct Urine LC-MS-MS Analysis for Screening of Novel Substances in Drug Abusers

Transcription

1 Application of Direct Urine LC-MS-MS Analysis for of Novel Substances in Drug Abusers Helena K. Nordgren 1, Per Holmgren 2,', Paula Liljeberg 3, Nadja Eriksson 3, and Olof Beck 1,t 1Department of Medicine, Division of Clinical Pharmacology, Karolinska Instituter, Sweden; 2National Board of Forensic Medicine, Department of Forensic Chemistry, University Hospital, Sweden; and 3Department of Neuroscience, Karolinska Instituter and University Hospital, Stockholm, Sweden Abstract A newly developed liquid chromatography-tandem mass spectrometry (LC-MS-MS) method was used to study 3000 human urine samples from 3 different populations for 23 analytes covering phenylethylamines, benzylpiperazine, and non-benzodiazepine hypnotics. Direct injection of urine and LC-MS-MS with rapid chromatography and atmospheric pressure chemical ionization was used in the screening step. The cutoff levels were chosen to be at the limit of detection for most analytes to identify as many positive samples as possible. Typically one ion transition was monitored from the pseudo-molecular ions in the multiple reaction monitoring mode. Of the 797 positive screening findings, 518 (65%) were confirmed by a second LC-MS-MS analysis including solid-phase extraction. Confirmed analytical findings included 22 cases positive for N-benzylpiperazine, 88 for 3,4-methylenedioxy-Nmethylamphetamine and metabolites, 4 for 1-phenyl-2-butylamine, 24 for zolpidem and metabolites, 118 for zopiclone and metabolites, and 1 for zaleplon. In conclusion, LC-MS-MS was found to be a robust alternative for drugs of abuse screening, offering high sensitivity compared with immunochemical screening methodology. Introduction Over the past decade, there have been several reports indicating misuse of new substances such as ecstasy and analogues thereof (1-4), y-hydroxybutyrate (GLIB) (5,6), non-benzodiazepine hypnotics (7-9), tryptamines (10), and N-benzylpiperazine (11). Most often, this information originates from cases of fatal intoxication and from seizures by police and customs. It, therefore, remains unknown how widespread their use is in the population, although there are indications that some of these new substances have become more widely used. For example, in one report GHB overdose represented 3.1% of all toxicological emergencies (12), and a survey from Wales in Great Britain revealed that 7.5% of the adolescents had ingested psilocybe mushroom (13). * Deceased. ~ Author to whom correspondence should be addesssed. Department of Clinical Pharmacology, L7:05, Karolinska University Hospital, S Stockholm, Sweden. olof.beck@karolinska.se. Drug testing is routinely used in clinical and forensic laboratories and the result from such testing represent one important source for estimating the prevalence of drugs of abuse in the society. There are several examples where prevalence data have been obtained by drug testing. In a U.S. study among 836 commercial truck drivers on duty, urine drug testing revealed that 9.5% were positive for amphetamines and cocaine, 4.3% for cannabis, and 1.6% for opiates (14). In a Norwegian traffic medicine study performed during one year, 3343 individuals were sampled upon suspicion that they were driving under the influence of drugs other than alcohol. The prevalence of drugs of abuse was 31% for benzodiazepines, 30% for tetrahydrocannabinol, 28% for amphetamine, and 8% for morphine (15). It would be valuable in such studies to complement also with data for newer substances. Immunochemical screening, which is the most common methodology used for screening large amounts of samples in urine drug testing, is limited to the more established substances, liowever, other techniques such as mass spectrometry (MS) can also be used to screen for new drugs of abuse with high sample throughput (16,17). Direct screening of urine by liquid chromatography-tandem MS (LC-MS-MS) is a promising method for screening large amounts of urine samples with high sensitivity and reliability (18) and proven to be a useful alternative to immunochemical screening analysis in the search for new drugs of abuse (19-21). In this study, a newly developed LC-MS-MS method for direct analysis in urine is applied in three different subpopulations to investigate the occurrence of phenylethylamines, N-benzylpiperazine and non-benzodiazepine hypnotics. A list of the abbreviations of the compounds analyzed is presented in Table I. Material and Methods Apparatus The LC-MS-MS system consisted of a vacuum degasser (Waters Milford, MA), a series 200 mixer, series 200 autosampler, two series 200 micro pumps (PerkinElmer, Norwalk, CT), a column oven kept at 40~ (Kontron, Zi~rich, Switzerland), and a Sciex 2000 triple-quadrupole MS (Applied Biosystems, MDS 234 Reproduction (photocopying) of editorial content of this journal is prohibited without publisher's permission.

2 Sciex, Concord, ON, Canada). The system was controlled via Analyst 1.1 software (PE Sciex). The mass spectrometer was connected to the LC via an atmospheric pressure chemical ionization interface kept at 450~ The interface was set to monitor positive ions in the multiple reaction monitoring mode. LC-MS-MS analysis The screening of urine included a 10-time dilution with ultra-pure water, rapid chromatography, and monitoring of ion-transitions from the pseudo-molecular ions, as described elsewhere (19). A slight modification was introduced, consisting of a 0.5-min prolonged plateau in the gradient elution to increase the robustness of the analysis. The total analysis time was 7 min. The analytes detected in the screening analysis were considered as preliminary positive results and subjected to a confirmation analysis. This step included a sample preparation with solid-phase extraction and a different system for chromatographic separation (19,22). In the confirmation of 3,4- methylenedioxy-n-methylamphetamine (MDMA), zopiclone, and zolpidem always included parent compund and metabolites in the confirmation analysis. The performance of the method was monitored by analyzing control samples at several levels. The detectability was measured as the percent of positive outcomes in the samples spiked at the cutoff levels (19), and it was calculated for 39 batches of samples. Approximately 90 urine samples could be analyzed in 24 h, including sample preparation and evaluation of results. The methodology was generally robust, but 7% of the batches stopped during overnight runs and had to be re-started in the morning. The precolumn was changed about every 200 samples, and the MS was cleaned every samples. Urine samples A total of 3000 urine samples were collected from three different subpopulations. Approval was obtained from the ethics committee of the Karolinska Institutet (Dnr KI ). The Maria Ungdom Adolescent Dependence Clinic. This clinic receives patients under 20 years of age from the greater Stockholm area. Only first time visitors were included in the study. The subjects had a median age of 17 years (range 12-21), and 54% were male. In total, 1000 urine samples were obtained from 941 individuals (a maximum of 4 samples per individual) during a 17-month period in The urine was stored in plastic tubes at -20~ for up to 3 months prior to analysis. The Stockholm Methadone Maintenance Treatment Program. A total of 1000 urine samples were collected from methadone patients. The subjects had a median age of 43 years (range 18-68), and 73% were males. The samples were randomly selected from the flow of samples taken for routine urine drug testing. Only urine samples containing > 10 ml were in- Table I. The Analytes, the Cutoff Levels, and Detectability of Standards at Cutoff Level Between Batches Ion Transitions Cutoff Level Detectability Compound Abbreviations (m/z) (ng/ml) (%) 3,4-Ethylenedioxyamphetamine 2-Methylamino-l-(3,4-methylenedioxyphenyl)-butane N-Ethyl-3,4-methylenedioxyamphetamine 3,4-Methylenedioxy-N-propylamphetamine 2,4,5-Trimethoxyamphetamine 3,4-Methylenedioxy-N-methylamphetamine MDMA-ds (internal standard) 3,4-Methylenedioxyamphetamine 4-Hydroxy-3-methoxy-N-methylamphetamine 2,5-Dimethoxy-4-bromoamphetamine 2,5-Dimethoxy-4-iodoamphetamine 1 -(2,5-Dimethoxyphenyl)-2-aminopropane 1 -(4-Bromo-2,5-dimethoxyphenyl)-2-ethaneamine N,N-DimethyI-MDA N,N-Dimethyl-l-phenyl-2-ethaneamine 1 -Phenyl-2-butylamine N-Benzylpiperazine Zopiclone nor-zopiclone Zopidone-N-oxide Zolpidem Zolpidem metabolite 1 Zolpidem metabolite 2 Zaleplon RP29481 (internal standard) EDA MBDB MDEA MDPA TMA-2 MDMA MDA HMMA DOB DOI 2,5-DMA 2-CB ~ ~ ) ) ~ ) ) ) ) ~ ) ) ~ ~ ) ~ ) ) ) ) ~ ~ ~ ) ) ~ )

3 cluded. The samples were from 322 individuals (a maximum of 11 samples per individual with at least 6 days in between samplings) and collected during a 5-month period in The samples were stored in plastic tubes for 1-2 days at 8~ and then for up to 4 months at -20~ prior to analysis. A forensic chemistry laboratory. A total of 1000 urine samples from individuals suspected for minor narcotics offence (according to the Penal Law on Narcotics in Sweden) were obtained from the National Board of Forensic Medicine, Link6ping. The subjects had a median age of 24 years (range 13-70), and 72% were males. The samples were randomly collected during a 6-month period in The samples had been stored at 4~ for up to 6 months, after which a 2.5-mL aliquot of the sample was transferred to a plastic tube and stored for 0-4 months at -20~ prior to analysis. Results and Discussion Application of LC-MS-MS for the detection of drugs of abuse in urine The detectability of the assay was determined to be % (Table I). Of the total of 3000 urine samples included in this study, 528 (17.9%) were found to be positive for any drug in the screening assay. In these samples 797 separate analytical Table II. s from the 3000 Studied that Screened Positive Number of Number of Positive s Confirmedin the Positive Analyte Analysis s* N-Benzylpiperazine CB 7 N,N-DimethyI-MDA 5 N,N-Dimethyl-l-phenyl-2-ethaneamine 4 2,5-DMA 4 DOI 11 MBDB 5 MDEA 8 MDMA, MDA, and/or HMMA* MDMA* MDA* HMMA* MDPA 13 1-Phenyl-2-butylamine 64 4 TMA-2 34 Zolpidem and metabolites* Zolpidem* 43 6 Zolpidem metabolite 1' Zolpidem metabolite 2* Zaleplon 18 1 Zopiclone and metabolites* Zopiclone* Norzopiclone* Zopiclone-N-oxide* * A positive screening analysis for any substance resulted in confirmation of all related substances (such as metabolites). 236 findings were made (Table II). All 23 analytes except 3,4- ethylenedioxyamphetamine (EDA) and 2,5-dimethoxy-4-bromoamphetamine (DOB) were detected. Of the 797 analytical findings, 628 (79%) were confirmed to be positive (Table II). Out of the 528 samples that were screened positive, 237 (45%) were confirmed to be positive and in nine of those there was evidence for the intake of more than one substance. ~vo criteria had to be fulfilled for a positive screening result: a correct relative retention time and a signal above the detection level. About one third of the findings that were positive in the screening analysis turned out to be negative in the confirmation analysis. For most analytes, there were rather few (0-13) false-positive results, but for N-benzylpiperazine, 1- phenyl-2-butylamine, TMA-2, zaleplon, and zolpidem, the rate was somewhat higher (up to 64). In order to fulfill the requirements of a screening method for clinical routine use, this high rate of false-positive results clearly reduces its usefulness. One example of a urine sample with a false-positive screening result for 1-phenyl-2-butylamine is shown in Figure 1. In the screening assay an interfering peak of high intensity coeluted with the analyte, but the result from the confirmation analysis demonstrated that it did not contain 1-phenyl-2-butylamine. A more typical example of a unconfirmed positive screening result is shown for N,N-dimethyl-MDA in Figure 2. One obvious way to reduce the incidence of false positives in the screening step is to raise the cutoff levels that, in this study, were close to the detection limit in most cases. For example, 16 urine samples generated falsepositive results for MDA in the screening anal- Prevalence ysis, but this number could be reduced to 4 if the cutoff for MDA were modified from 100 to 200 ng/ml without any loss of true-positive resuits. However, increased cutoff levels result in o.4 loss of sensitivity (18) and may not always be the best solution. Another possibility to reduce the number of false positives is to use a more selective chromatographic system with a higher separation power. This would, however, lead to longer analysis times and a reduced capacity. Still another way to increase selec- 2.9 tivity is to monitor more than one ion transition for each analyte. This possibility was examined for DOB, for which the highest rate of false-positive results was observed. By including monitoring of two additional ion tran- 0.1 sitions, no false-positive results remained (see example in Figure 3). For DOB, this interfer- 0.8 ence was observed throughout the entire study, whereas, for other analytes, interfering peaks appeared more sporadically. In the confirmation analysis, a higher Based on Confirmation Results (%) o.o3 3.9 selectivity was introduced by the sample preparation by SPE and the changed chromatographic system. This also resulted in an increased analytical sensitivity, as compared to the screening assay. For all analytes, the same ion transition(s) was monitored in the confirmation and the screening. Transitions

4 Standard ~:, 250 =~ 200 i 'Y 0,ei ,v ,t..- x 800 Standard [ 0oo 400 2O0 o 0 A $ ~ Confirmation ~91.0 B C !2 o ~0 ~ Confirmation D $ > Figure 1. Screened authentic urine sample containing ng/ml 1- phenyl-2-butylamine (A). Confirmation was negative (B). Screened spiked urine sample containing 500 ng/ml 1-phenyl-2-butylamine (C). Confirmed spiked urine sample containing 5000 ng/ml. The arrow indicates the correct retention time of the analyte (D). 7 i ' 3 j= 1 o, S 6 Confirmation B ~ 5o t ~ i 3O=o o Standard 15 A ,4 ] J" 208.1"->'162.9 I o....,~-r ~,','~ ~ 5 0 Confirmation C D Timelmln) Figure 2. Screened authentic urine sample containing 58 ng/ml N,NdimethyI-MDA (A). Confirmation was negative (g). Screened spiked urine sample containing 50 ng/ml N,N-dimethyI-MDA (C). Confirmed spiked urine sample containing 5 ng/ml (D). The arrow indicates the correct retention time of the analyte. Patient sample ~. 20 x 15 l,o = s,v Minutes Standard > I0 is t $ ~ I 2 A D t Is $ >229.0 Is 2 10 ~ _ ~ 10 5 s $ Or I I 2 3 Minutes Minutes B 274.o.229.o E j s. 10 C ~ a $ I I Minutes Minutes Minutes Figure 3. These chromatograms demonstrate the benefit of using more than one ion-transition in the screening analysis to avoid false-positive results. The use of the third ion-transition (274.2 ~ 178.3) determines that this sample did not contain DOB. Authentic urine sample determined to contain 431 ng/ml of DOB when using only one ion-transition (274.2 ~ 257.0)(A-C). Spiked urine sample containing 100 ng/ml DOB. The arrow indicates the correct retention time of the anal,fie (D-F). 2 ' L F I 237

5 Table III. Findings in the 3000 s Adolescent Forensic Methadone s s s N-Benzylpiperazine 4 7 MDMA/MDA/HMMA Phenyl-2-butylamine 4 Zolpidem 2 17 Zaleplon Zopiclone monitored for all substances analyzed within this study are listed in Table I. To secure the accuracy of the identification of analytes, it was recently proposed to use a minimum of two ion transitions in LC-MS-MS (23,24). The experience from this work is that the selectivity obtained in sample preparation and chromatography is also important and should be considered. The findings in the three subpopulations are summarized in Table III. 4-Hydroxy-3-methoxy-N-methylamphetamine (HMMA) could be detected in 16% of the 88 samples positive for MDMA and metabolites, zopiclone-n-oxide in 45% of the 118 samples positive for zopiclone and metabolites, and zolpidem in 38% of the 24 samples positive for zolpidem and metabolites. If HMMA, zopiclone-n-oxide, and zotpidem had been excluded from the screening analysis, only one sample positive for zolpidem and metabolites would have been missed. These results suggest that HMMA, zopiclone-n-oxide, and possibly zolpidem could be excluded from the screening analysis with only a minor effect on the results. Zaleplon is mainly excreted in urine as metabolites, with less than 1% as zaleplon (25). If the zaleplon metabolites had been included in this study, more samples positive for zaleplon metabolites might have been detected. Unfortunately, the metabolites were not available. The highest rate of positive samples was observed in the samples obtained from the forensic cases. This is not surprising as the samples are collected on the basis of reasonable suspicion of being under the influence of drugs, and the rate of detecting abused drugs in these cases is ~85%. The rate of positive result for abused drugs in the clinical samples is much lower and in the order of 10-15%. The findings in the samples from the adolescent clinic resembled those from the forensic cases with the difference of a lower incidence of positive results and the lack of 1-phenyl-2-butylamine. There were differences between the subpopulations regarding the frequency of zopicione and zolpidem findings. Only hypnotics were present among the methadone patients. None of the samples from methadone patients were positive for phenylethylamines and N-benzylpiperazine. This is in accordance with the clinical conception that methadone patients rarely are interested in using this type of "new" drugs. The occurrence of zopiclone and N-benzylpiperazine among drug abusers have been indicated in more limited clinical and forensic investigations (7-9,11,22). The four samples positive for 1-phenyl-2-butylamine could not be further confirmed with LC-MS-MS by using additional ion-transitions because of the lack of other ion-transitions. Instead, GC-MS was used to secure the identity of the findings. 1-Phenyl-2-butylamine has been shown to be a CNS stimulant (26) and is under legislation as a narcotic in Sweden. Still, there is very little available information about this drug, and the origin for 1- phenyl-2-butylamine in the four samples has to be considered to be unknown and will be subject to further investigation. Conclusions Direct screening of urine by LC-MS-MS is a useful technique for detection of drugs of abuse in large volumes of samples, and it is especially valuable for novel substances for which no immunochemical method is available. Using LC-MS-MS for direct screening makes it easy to tailor the menu of the analyzed substances to match new trends in comparison to more dedicated analytical methods. With this technique, we were able to study the prevalence of 23 analytes in three different populations. In future work, an increased selectivity by monitoring at least two ion transitions is recommended. References 1. European Monitoring Centre for Drugs and Drug Addiction (EM- CDDA) Annual Report on the State of the Drug Problem in the European Union and Norway, Lisbon, Portugal, S.P. Elliott. Fatal poisoning with a new phenylethylamine: 4-methylthioamphetamine (4-MTA). J. Anal. Toxicol. 24:85-89 (2000). 3. C. Giroud, M. Augsburger, L. Rivier, P. Mangin, F. Sadeghipour, E. Varesio, J.L. Veuthey, and P. Kamalaprija. 2C-B: a new psychoactive phenylethylamine recently discovered in ecstasy tablets sold on the Swiss black market. J. Anal. Toxicol. 22' (1998). 4. A.S. Christophersen. Amphetamine designer drugs--an overview and epidemiology. Toxicol. Lett : (2000). 5. P.C. Kam and F.F. Yoong. Gamma-hydroxybutyric acid: an emerging recreational drug. Anaesthesia 53: (1998). 6. H.K. Louagie, A.G. Verstraete, C.J. De Soete, D.G. Baetens, and P.A. Calle. A sudden awakening from a near coma after combined intake of gamma-hydroxybutyric acid (GHB) and ethanol. Clin. Toxicol. 35: (1997). 7. K. Aranko, M. Henriksson, C. Hublin, and A.M. Seppalainen. Misuse of zopicione and convulsions during withdrawal. Pharmacopsychiatry 24: (1991). 8. S. Sikdar and S.M. Ruben. Zopiclone abuse among polydrug users. Addiction 91: (1996). 9. S. Rooney and J.J. O'Conner. Zopiclone, a current drug of misuse. Addiction 93:925 (1998). 10. S. Byqvist. CANs rapporteringssystem om droger (CRD)--Tendenser under v~ren CentralfOrbundet f6r alkohol- och narkotikaupplysning, Stockholm, Sweden, 2003, p M. Wikstr~m, P. Holmgren, and J. Ahlner. A2 (N-benzylpiperazine) a new drug of abuse in Sweden. J. Anal. Toxicol. 28:67-70 (2004). 12. O. Miro. S. Nogue, G. Espinosa, J. To-Figueras, and M. Sanchez. Trends in illicit drug emergencies: the emerging role of gammahydroxybutyrate. J. Toxicol. Clin. Toxicol. 40: (2002). 13. C. Smith and D. Nutbeam. Adolescent drug use in Wales. Br. J. Addiction 87: (1992). 14. F.J. Couper, M. Pemberton, A. Jarvis, M. Hughes, and B.K. Logan. Prevalence of drug use in commercial tractor-trailer drivers. J. Forensic Sci. 47: (2002). 15. S. Skurtveit, B. Abotnes, and A.S. Christophersen. Drugged drivers 238

6 in Norway with benzodiazepine detections. Forensic Sci. Int. 125: (20O2). 16. J. Zweigenbaum, K. Heinig, S. Steinborner, T. Wachs, and J. Henion. High-throughput bioanalytical LC/MS/MS determination of benzodiazepines in human urine: 1000 samples per 12 hours. Anal Chem. 71: (1999). 17. W. Weinmann and M. Svoboda. Fast screening for drugs of abuse by solid-phase extraction combined with flow-injection ionspraytandem mass spectrometry. J. Anal. Toxicol. 22" (1998). 18. H.K. Nordgren and O. Beck. Direct screeing of urine for MDMA and MDA by liquid chromatography-tandem mass spectrometry. J. Anal ToxicoL 27" (2003). 19. H.K. Nordgren and O. Beck. Multicomponent screening for drugs of abuse: direct analysis of urine by LC-MS-MS. Ther. Drug Monit. 26:90-97 (2004). 20. S.R. Needham, P.M. Jeanville, P.R. Brown, and E.S. Estape. Performance of a pentafluorophenylpropyl stationary phase for the electrospray ionization high-performance liquid chromatography-mass spectrometry assay of cocaine and its metabolite ecgonine methyl ester in human urine. J. Chromatogr. B 748:77-87 (2OO0). 21. R. Dams, C.M. Murphy, W.E. Lambert, and M.A. Huestis. Urine drug testing for opioids, cocaine and metabolites by direct injection liquid chromatography/tandem mass spectrometry. Rapid Commun. Mass Spectrom. 17" (2003). 22. H.K. Nordgren, K. Bodin, and O. Beck. Chromatographic screening for zopiclone and metabolites in urine using liquid chromatography and liquid chromatography-mass spectrometry techniques. Ther. Drug Monit. 24: (2002). 23. L. Rivier. Criteria for the identification of compounds by liquid chromatography-mass spectrometry and liquid chromatography-multiple mass spectrometry in forensic toxicology and doping analysis. Anal Chim. Acta 492" (2003). 24. European Union Decision 2002/657/EC Off. J. Eur. Communities (2002). 25. A.S. Rosen, P. Fourni~, M. Darwish, P. Danjou, and S.M. Troy. Zaleplon pharmacokinetics and absolute bioavailability. Biopharm. Drug Dispos. 20" (1999). 26. D.E Marsh. The pharmacological activity of 1-phenyl-2-butylamine. J. PharmacoL Exp. Ther. 94: (1948). Manuscript received March 5, 2004; revision received August 11,