Ultrafast Screening of Synthetic Cannabinoids and Synthetic Cathinones in Urine by RapidFire-Tandem Mass Spectrometry

Transcription

1 Journal of Analytical Toxicology, 2016;40: doi: /jat/bkw025 Technical Note Technical Note Ultrafast Screening of Synthetic Cannabinoids and Synthetic Cathinones in Urine by RapidFire-Tandem Mass Spectrometry Jillian R. Neifeld*, Laura E. Regester, Justin M. Holler, Shawn P. Vorce, Joseph Magluilo Jr, Gerardo Ramos, and Thomas Z. Bosy Division of Forensic Toxicology, Armed Forces Medical Examiner System, 115 Purple Heart Drive, Dover, DE 19902, USA *Author to whom correspondence should be addressed. Abstract Screening for emerging drugs of abuse, specifically synthetic cathinones and synthetic cannabinoids, is difficult for high-throughput laboratories as immunoassay kits are often unavailable. Consequently, most laboratories employ liquid chromatography tandem mass spectrometry (LC MS-MS) screening, which can be complex and time consuming as these techniques may require involved sample preparation and lengthy analysis times. The increasing demand for novel psychoactive substance testing necessitates alternative screening methods that are sensitive, fast and versatile. The RapidFire tandem mass spectrometry system (RF MS-MS) provides a rapid and highly specific screen for these emerging drugs of abuse with minimal sample preparation and an instrumental analysis time of <14 s per sample. Presented here are two RF MS-MS screening methods used to analyze 28 emerging drugs of abuse, 14 synthetic cannabinoids and 14 synthetic cathinones, in urine with run times of 9 and 12.6 s, respectively. Sample preparation and hydrolysis were performed in a 96-well plate with one multiple reaction monitoring transition used for the identification of each compound. Eighteen thousand urine specimens were screened by liquid liquid extraction followed by LC MS-MS analysis, and the results were compared with those obtained using the RF MS- MS screening method. The analytical data illustrate the advantages of the RF MS-MS methods. Introduction Designer drug use has become more prevalent over the last several years, as indicated by the increased number of law enforcement seizures and emergency room visits (1, 2). Toxicology laboratories have had to develop and adjust their testing panel and paradigms to meet the increasing demand for analysis of these novel psychoactive substances (NPSs). The latest trend for designer drugs has been dominated by two classes of compounds: synthetic cathinones/designer amphetamines and synthetic cannabinoids (3). Synthetic cathinones/designer amphetamines have emerged as popular drugs of abuse. These legal highs are referred to as bath salts and tend to produce amphetamine-like symptoms when ingested. Synthetic cathinones from several different structural groups have been detected in these legal high products; they include the 2C series (2,5-dimethoxyphenethylamines), α-pyrrolidinopropiophenones, β-keto amphetamines, piperazines, substituted amines and PCP analogs (4, 5). Currently, there are limited commercially available immunoassay (IA) screening kits capable of detecting synthetic cathinones. As a result, toxicology laboratories have been forced to develop alternative screening technologies and processes. Several studies have been published reporting the cross-reactivity of designer amphetamines with various commercial (6 8) and in-house (9) amphetamine and methamphetamine IA kits. While some bath salts produce positive IA screening results, many more do not, which means that these compounds would remain undetected. The compounds generating a positive IA screen would most likely be confirmed using a traditional amphetamine/methamphetamine confirmation method; this would result in numerous compounds going unnoticed by current techniques. Synthetic cathinone structures are also frequently modified, which makes them increasingly difficult to detect using current IA screening methods (10 12). Consequently, existing The Author Published by Oxford University Press. All rights reserved. For Permissions, please journals.permissions@oup.com 379

2 380 Neifeld et al. confirmation methodologies, such as full scan gas chromatography/mass spectrometry (GC MS) (5), liquid chromatography tandem mass spectrometry (LC MS-MS) (13) and time-of-flight mass spectrometry (ToF MS) (3), are being used to replace the IA screen. Unfortunately, these methods are laborious and require extensive data analysis and technical expertise. Synthetic cannabinoids are also referred to as legal highs, even though many of these compounds have been scheduled by the Drug Enforcement Agency. Commonly known as Spice or K2, these products have similar pharmacological effects to those of tetrahydrocannabinol (THC), the psychoactive compound found in marijuana. As with synthetic cathinones, the structures of these compounds are modified frequently (14). As current compounds are controlled or scheduled by government agencies, the manufacturers simply alter the structure. The availability of commercial IA kits capable of recognizing synthetic cannabinoids is limited as, typically, IA kits were developed to identify specific structural classes, which means that these products are unlikely to detect more than one or two spice compounds per kit. To expand IA detection capability, kits designed to identify different structural moieties may be used (15). Although some synthetic cannabinoids are structurally similar to THC, they do not cross-react with IA kits designed for common THC metabolites (1, 16). Many labs use LC MS-MS methods to screen for synthetic cannabinoids. The lengthy extraction involved with using these methods requires a large amount of sample (1 3 ml, depending on the extraction) and significant volumes of solvent waste. Typical run times for LC MS-MS analyses are 8 10 min per injection with a mobile phase flow rate of ml/min (14, 17). New synthetic cannabinoids are identified frequently, and updating and validating existing LC MS-MS methods is a challenge both technically and financially for toxicology laboratories. Presented here are two RapidFire tandem mass spectrometry system (RF MS-MS) screening methods used to analyze 28 emerging drugs of abuse (14 synthetic cathinones and 14 synthetic cannabinoids) in urine with run times of 9 and 12.6 s, respectively. Method effectiveness was evaluated by analyzing 18,000 randomly collected urine samples for both designer drug classes. The method was fully validated prior to use and analysis of the 18,000 samples. Samples were previously confirmed for synthetic cathinones using an existing LC MS-MS method. Experimental Chemicals and reagents All solvents were purchased from Fisher Scientific (Pittsburgh, PA). Elution solvents were of Optima grade, and remaining solvents were of HPLC grade. Synthetic cannabinoids, synthetic cathinones and internal standards JWH 073 N-CO-d5 and MDPV-d8 were purchased from Cerilliant (Round Rock, TX) or Cayman Chemical (Ann Arbor, MI). Sodium hydroxide (5 N Na) was purchased from Aldrich (Milwaukee, WI). Ammonium acetate was purchased from Mallinckrodt Chemicals (St. Louis, MO). RapidFire A2 cartridges were purchased from Agilent (Palo Alto, CA). Standard 96 well U -bottom plates were purchased from Greiner bio-one (Unionville, NC). K2 (Synthetic Cannabinoids-1) Urine Enzyme IA kits were purchased from Immunalysis (Pomona, CA). Sample preparation and extraction Urine samples (50 µl) were placed in individual wells of a 96-well plate. An internal standard (25 µl at a 1 µg/ml concentration) and 0.3 N Na (25 µl for hydrolysis of the synthetic cannabinoid metabolites) wereaddedtoeachwell,mixed,andtheplateswereincubatedatroom temperature for 15 min. Following incubation, 150 µl of diluent (50:50 deionized water methanol with 0.5% formic acid) was added. Each plate was sealed with a clear sealant using the Agilent PlateLoc sealer, centrifuged and refrigerated until time of analysis. Instrumental analysis Urine samples were analyzed for synthetic cannabinoids and synthetic cathinones using an Agilent RapidFire 365 coupled to an Agilent 6460 tandem mass spectrometer (Palo Alto, CA). RapidFire 365 The flow scheme of the RapidFire is shown in Table I, a schematic representation for each valve state can be found at com/cs/library/usermanuals/public/g _rapidfire365_ User.pdf on pages A standard 96-well plate was used; each plate contained 1 calibrator (5 ng/ml for all synthetic cannabinoids and 10 ng/ml for all synthetic cathinones), 3 controls (one low control at 75% of the cutoff concentration, one high control at 125% of the cutoff concentration and one negative control) and 92 samples. RapidFire parameters for both synthetic cannabinoids and synthetic cathinones are shown in Table II. The injection volume changes Table I. RapidFire Flow Path State #1 State #2 State #3 State #4 State #5 State #6 Table II. RapidFire Parameters SynCan Buffer A Buffer B Buffer C Injection volume SPE cartridge RF State #1 RF State #2 RF State #3 RF State #4 RF State #5 Bath Salts Buffer A Buffer B Buffer C Injection volume SPE cartridge RF State #1 RF State #2 RF State #3 RF State #4 RF State #5 Aspirate/sip sample Load sample/wash cartridge Extra wash of cartridge Elute sample Re-equilibrate system Flush sipper (at end of plate or when clog is in system) Description 10 mm ammonium acetate w/0.1% formic acid (1.5 ml/min) 50:50 Me:water w/0.1% formic acid (2.0 ml/min) 85:15 ethyl acetate isopropanol (1.25 ml/min) ml A2 (C4) (Agilent) 600 ms 1,500 ms 5,000 ms 5,000 ms 500 ms 10 mm ammonium acetate w/0.1% formic acid (1.5 ml/min) Same as C (1.25 ml/min) 50:25:25 water acetone acetonitrile w/0.09% formic acid and 0.01% trifluoroacetic acid (1.25 ml/min) ml A2 (C4) (Agilent) 600 ms 3,000 ms None 5,000 ms 500 ms

3 RapidFire/MS-MS Screen 381 Table III. MS-MS Parameters Description Synthetic cannabinoids Gas temperature 300 C Gas flow 13 L/min Nebulizer 50 psi Sheath gas 300 C Sheath gas flow 11 L/min Nozzle voltage 1,000 V Time filter (s) 0.02 Dwell time 15 ms Capillary voltage 3,000 V Synthetic cathinones Gas temperature 300 C Gas flow 11 L/min Nebulizer 50 psi Sheath gas 300 C Sheath gas flow 11 L/min Nozzle voltage 500 V Time filter (s) 0.02 Dwell time 10 ms Capillary voltage 3000 V Table IV. Synthetic Cathinone Panel depending on the size of the sample loop. During this study, the loop was cut so that µl of sample was sipped and 10 µl of each sample was used for analysis. The total run times for the synthetic cathinones and cannabinoids analyses were 9.1 and 12.6 s, respectively. The RapidFire has four plate stacks. One stack remains empty for completed plates (the waste stack ). The other three stacks can hold a maximum of 21 plates each, or 63 plates total. Each plate holds 92 samples, which means that 5,796 samples can be analyzed in one run. A robotic arm called the bench bot moves the plates from the stacks to the stage and back. If sample tracking is required, a barcode scanner is available that can scan each plate before the plate is analyzed. Mass spectrometer Both the synthetic cathinone and the synthetic cannabinoid methods used electrospray ionization. Table III shows the mass spectrometry source methods for both compound classes. A single multiple reaction monitoring (MRM) transition was monitored for each compound. The synthetic cathinone transitions are shown in Table IV, and the synthetic cannabinoid transitions are shown in Table V. Analyte Acronym Q1 Q3 Frag (V) CE 1 2-(4-Iodo-2,5-dimethoxyphenyl)-N-[(2-methoxyphenyl)methyl]ethanamine 25I-NBOMe ,4-Methylenedioxy-α-pyrrolidinovalerophenone MDPV (3,4-Methylenedioxyphenyl)-2-methylamino-1-pentanone Pentylone α-pyrrolidinovalerophenone α-pvp (3,4-Methylenedioxyphenyl)-2-methylamino-1-butanone Butylone (3,4-Methylenedioxyphenyl)-2-ethylamino-1-propanone Ethylone α-pyrrolidinobutyrophenone α-pbp (3,4-Methylenedioxyphenyl)-2-methylamino-1-propanone Methylone α-pyrrolidinopropiophenone α-ppp Methyl-N-ethylcathinone 4-MEC Methylmethcathinone Mephedrone N-Ethylcathinone NEC (Methylamino)-1-phenylbutan-1-one Buphedrone (2-Aminopropyl)benzofuran 5-APB D 8-3,4-Methylenedioxy-α-pyrrolidinovalerophenone D 8 -MDPV Table V. Synthetic Cannabinoid Panel Analyte Acronym Q1 Q3 Frag (V) CE 1 JWH-018 N-pentanoic acid JWH-018 N-CO JWH-018 N-(5-hydroxypentyl) JWH-018 N JWH-073 N-butanoic acid JWH-073 N-CO JWH-073 N-(4-hydroxybutyl) JWH-073 N JWH-250 N-(5-carboxypentyl) JWH-250 N-CO JWH-250 N-(5-hydroxypentyl) JWH-250 N JWH-081 N-(5-hydroxypentyl) JWH-081 N JWH-122 N-(5-hydroxypentyl) JWH-122 N AM2201 N-(4-hydroxypentyl) AM2201 N MAM2201 N-pentanoic acid MAM2201 N-CO RCS-4 N-(5-carboxypentyl) RCS-4 N-CO UR-144 N-pentanoic acid UR-144 N-CO XLR11 N-(4-hydroxypentyl) XLR11 N AKB48 N-pentanoic acid AKB48 N-CO D 5 -JWH-073 N-butanoic acid D 5 -JWH-073 N-CO

4 382 Neifeld et al. Method validation Linearity Synthetic cannabinoids demonstrated linearity from 1.00 to 100 ng/ml with a limit of detection (LOD)/limit of quantitation (LOQ) of 1.00 ng/ml. Synthetic cathinones demonstrated linearity from 5.00 to 100 ng/ml with an LOD/LOQ of 5.00 ng/ml. All compounds had correlation coefficients (R 2 ) above Selectivity and specificity Ten drug-free urine samples from separate sources were analyzed for matrix interferences. No matrix interferences were observed for the synthetic cathinones. For the synthetic cannabinoids analysis, several of the negative urines produced quantifiable signal responses for the primary MRM transition ( ) for JWH 073 CO. Because of this, a secondary MRM transition ( ) was used. Other compounds (172 structurally similar and nonsimilar compounds shown in Table VI) were analyzed for possible interferences from either co-eluting compounds or matching MRM ion transitions. The compounds were fortified at 1,000 ng/ml into both negative urine samples and positive low control samples. None of the 172 compounds analyzed caused interferences in either the negative urine or the low control samples. Precision and accuracy Precision was calculated for all compounds at three concentrations: low (75% of the calibrator concentration), mid (calibrator concentration) and high (125% of the calibrator concentration). For the synthetic cathinones, the analyzed concentrations were 7.50, 10.0 and 12.5 ng/ml. For synthetic cannabinoids, the analyzed concentrations were 3.75, 5.00 and 6.25 ng/ml. Twelve aliquots of urine at the above concentrations were analyzed in one batch. The average concentration and coefficient of variation (CV) are illustrated in Tables VII and VIII for the synthetic cathinones and cannabinoids, respectively. All synthetic cathinones had CV values within 15%, except for 5-APB (low) and methylone (low and mid), which were within 25%. The synthetic cannabinoids all had CV values within 15% except for AM 2201 (low and mid), JWH 073 CO/JWH 018 (low) and XLR 11 (high), which were all within 25%. Accuracy was determined by comparing the mean to the theoretical concentration; this is represented as the percent difference (Table VII for synthetic cathinones and Table VIII for synthetic cannabinoids). For synthetic cathinones, accuracy calculations for each of the three concentrations were within 20%, except for methylone (all), α-pbp (high) and 4-methylethcathinone (mid), which were all within 30%. For synthetic cannabinoids, accuracy calculations were within 20%, except for JWH 073 CO/JWH 018 (all), JWH 018 CO (all) and RCS 4 CO (low and high), which were all within 30% of the theoretical concentration. Carryover Carryover was determined by measuring the analyte signal in a negative urine sample analyzed immediately after samples with analyte concentrations of 200 ng/ml (2 ULOL) and 500 ng/ml (5 ULOL) were injected. Carryover was not observed for any synthetic cathinones at 500 ng/ml. Carryover was observed for the synthetic cannabinoids MAM 2201 CO, AM 2201, JWH 250 CO, JWH 250, JWH 073 CO/JWH 018 and RCS 4 CO at 200 ng/ml (2 ULOL). For the remaining synthetic cannabinoids, carryover did not occur at 200 ng/ml but did occur at 500 ng/ml. Results Figure 1a is an example of the raw data for a single plate analyzed using RF MS-MS. Each individual peak is a separate sample injection; within each peak, 14 specific MRM transitions are monitored. Once all 96 samples are injected and the plate sequence is complete, the RF software will cut the one data file into individual data files each with their own distinct data file name. Each 96-well plate contained a calibrator, 3 controls and 92 samples with a 20-min analysis time per method per plate. This resulted in 92 samples, with controls and washes, being screened for 28 emerging drugs in approximately 45 min. Figure 1b shows JWH 073-N-CO-d5 only for each injection of the batch. Figure 1c shows JWH 018-N-CO only. In this batch, there were two presumptive positive samples, which are labeled. Figure 1d shows the MRM transition for JWH 018 N-CO from cut files for the batch. Specimen analysis The RF MS-MS was evaluated by analyzing 18,000 specimens that were previously extracted and analyzed for synthetic cathinones by traditional LC MS-MS. Analysis using the LC MS-MS method included 21 compounds, while the RF MS-MS method included 14 compounds. (The seven compounds not in the RF MS-MS method yielded no positive samples during the LC MS-MS screening process.) Original analysis was performed in MRM mode but did not include transitions for synthetic cannabinoids (presumptive positive synthetic cannabinoid samples from the RF MS-MS were confirmed using a LC MS-MS method). Table IX shows a comparison of both synthetic cathinone methods to include sample volume and preparation, analysis time and project completion time. The LC MS-MS method for synthetic cathinones identified six confirmed positive results while the RF MS-MS screening method identified 390 presumptive positive samples; the six confirmed positives were identified by the screening method. The synthetic cannabinoid analysis resulted in 76 presumptive positives by RF MS-MS with 37 confirmed by LC MS-MS for a 48.7% confirmation rate. Additionally, confirmed positive (LC MS-MS) synthetic cannabinoid specimens from a previous study were also analyzed by RF MS-MS and IA for comparison. Analysis using the LC MS-MS method included 13 compounds while the RF MS-MS method included 14 compounds. When using LC MS-MS, 261 samples confirmed positive using a 0.05 ng/ml cutoff concentration. Using a RF MS-MS method, 175 samples screened positive using a 5.00 ng/ml cutoff concentration (67% positive rate). Of the 86 samples that screened negative, 27 samples had confirmed concentrations of ng/ml (31%) and 57 had concentrations of <1.00 ng/ml (66%). Using an IA screening method, 120 samples screened positive for synthetic cannabinoid metabolites at a 10.0 ng/ml cutoff concentration (45% positive rate). Discussion The RapidFire system eliminates the typical chromatographic component expected during drug screening by MS, and, therefore, no separation of compounds occurs during analysis. This can be viewed as problematic if the analysis includes isobaric compounds such as ethylone and butylone or JWH 018 and JWH 073 CO. Because the assay is used as a screening technique in this laboratory, it was viewed as a nonissue. Hydrolysis Several different hydrolysis methods were evaluated to determine which would best hydrolyze the synthetic cannabinoids. Enzymatic

5 Table VI. Compounds Analyzed for Interferences 1,3-Dimethylamylamine (DMAA) Citalopram Lamotrigine Phencyclidine 11-nor-D9-THC-9-CO Clomipramine Lidocaine Pheniramine 1-Hydroxymidazolam Clonazepam Lisinopril Phenmetrazine 1-Hydroxytriazolam Clozapine Lorazepam Phenobarbital 3,4-Methylenedioxyamphetamine (MDA) Cocaine Loxapine Phentermine 3,4-Methylenedioxymethamphetamine (MDMA) Codeine Lysergic acid diethylamide (LSD) Phenylpropanolamine 3,4-Methylenedioxy-N-ethylamphetamine (MDEA) Cyclobenzaprine MAM2201 N-(4-hydroxypentyl) metabolite Phenytoin 3,4 Methylenedioxypyrovalerone (MDPV) D9-THC MAM2201 N-pentanoic acid metabolite PPP (α-pyrrolidinopropiophenone) 4-MEC (4-methylethcathinone) Desalkylflurazepam Maprotiline Procainamide 4-Methylmethcathinone (mephedrone) Dextromethorphan Meclizine Procaine 5F PB 22 Carboxyindole Diazepam Mefloquine Promethazine 6-Acetylmorphine Digoxin Meperidine Propoxyphene 7-Aminoclonazepam Dihydrocodeine Meprobamate Propranolol Acetominophen Diltiazem Methadone Pseudoephedrine ADB PINACA CO Diphenhydramine Methamphetamine Psilocin ADBICA CO Doxepin Methaqualone PVP (α-pyrrolidinovalerophenone) α-hydroxyalprazolam Doxylamine Methedrone (4-methoxymethcathinone) Pyrovalerone AKB48 N-CO Duloxetine Methylone (3,4-methylenedioxy N-methyl-cathinone) Quetiapine α-hydroxybutyrate Ephedrine Methylphenidate Quinine Alprazolam Eszopiclone Metoclopramide RCS-4 N-(5-carboxypentyl) AM2201 N-(4-hydroxypentyl) metabolite Ethylone Metoprolol RCS-4 N-(5-hydroxypentyl) Amitriptyline Etomidate Midazolam Salicylate Amobarbital Fentanyl Mirtazapine Secobarbital Amoxapine Flephedrone (4-fluoroephedrone) Morphine Sertraline Amphetamine Fluoxetine Morphine-3-glucuronide Strychnine Atenolol Fluvoxamine N-Acetyl Procainamide Temazepam Atropine Gamma-hydroxybutyric acid (GHB) Naphyrone Thioridazine Azacyclonol Gentamicin Nefazodone Tobramycin Benzoylecogonine Haloperidol N-Ethylcathinone Tramadol Benzphetamine Hydrocodone Norcodeine Trazodone Benztropine Hydromorphone Nordiazepam UR 144 N-(5-hydroxypentyl) metabolite Bromo-Diphenhydramine Hydroxyzine Norpropoxyphene UR 144 N-pentanoic acid metabolite Brompheniramine Imipramine Norsertraline Valproic acid Buphedrone JWH 018 N-(5-hydroxypentyl) metabolite Nortriptyline Vancomycin Bupivicaine JWH 018 N-pentanoic acid metabolite Norvenlafaxine Venlafaxine Buprenorphine JWH 073 N-(4-hydroxybutyl) metabolite Olanzapine Verapamil Bupropion JWH 073 N-butanoic acid metabolite Orphenadrine Warfarin Butalbital JWH 081 N-(5-hydroxypentyl) Oxazepam XLR11 N-(4-hydroxypentyl) metabolite Butylone JWH 122-N-(5-hydroxypentyl) metabolite Oxycodone Zaleplon Carbamazepine JWH 210 N-(5-carboxypentyl) Oxymorphone Zolpidem Carisoprodol JWH 210 N-(5-hydroxypentyl) Paroxetine Chlorpheniramine JWH 250 N-(5-carboxypentyl) PB 22 Carboxyindole Chlorpromazine JWH 250 N-(5-hydroxypentyl) Pentazocine Cimetidine Ketamine Pentobarbital RapidFire/MS-MS Screen 383

6 Table VII. Synthetic Cathinone Precision and Accuracy N = 12 Methylone Mephedrone 4-MEC PVP Butylone/ ethylone PBP 25I-NBoMe MDPV Pentylone PPP 5-APB NEC Buphedrone Cutoff Mean (ng/ml) CV 20.3% 15.0% 12.9% 11.1% 11.1% 6.7% 4.5% 6.5% 11.3% 14.1% 15.5% 14.6% 11.6% %diff 29.8% 8.9% 23.5% 8.5% 5.4% 19.8% 14.9% 14.7% 5.8% 8.0% 6.5% 13.2% 6.4% 75% cutoff Mean (ng/ml) CV 17.1% 12.9% 12.1% 10.2% 13.5% 11.6% 4.7% 9.0% 13.1% 8.9% 22.8% 10.3% 13.8% %diff 21.3% 11.6% 19.8% 16.2% 0.6% 18.8% 19.3% 17.8% 10.1% 11.5% 15.5% 14.9% 5.3% 125% cutoff Mean (ng/ml) CV 13.4% 13.5% 12.5% 6.0% 11.3% 12.1% 6.0% 6.3% 11.0% 9.4% 14.4% 11.2% 14.2% %diff 23.5% 6.3% 19.3% 7.6% 6.3% 20.9% 12.2% 14.9% 2.5% 5.4% 0.7% 14.5% 4.5% Table VIII. Synthetic Cannabinoid Precision and Accuracy N = 12 MAM2201 N- CO JWH-018 N-CO JWH 073 N- CO/JWH 018 JWH-250 N- RCS 4 N- CO XLR11 JWH-073 UR 144 CO AKB48 CO JWH-081 AM2201 JWH-122 Cutoff Mean (ng/ml) CV 11.3% 11.0% 13.5% 9.0% 8.1% 7.9% 14.2% 7.9% 6.5% 7.3% 24.4% 7.3% 10.1% %diff 12.7% 23.8% 25.4% 11.9% 19.7% 12.5% 3.4% 9.0% 11.4% 18.3% 8.7% 11.5% 13.8% 75% cutoff Mean (ng/ml) CV 10.9% 12.1% 16.6% 10.0% 15.7% 15.6% 12.6% 14.0% 12.4% 12.6% 24.2% 14.4% 11.0% %diff 11.9% 26.5% 29.6% 12.2% 26.5% 12.0% 12.0% 7.7% 9.3% 18.0% 1.8% 11.8% 10.9% 125% cutoff Mean (ng/ml) CV 11.5% 13.9% 11.8% 10.8% 14.5% 16.5% 10.0% 9.9% 7.9% 9.4% 13.9% 10.0% 11.7% %diff 12.5% 24.6% 22.4% 11.6% 22.6% 8.3% 7.7% 8.7% 9.7% 15.2% 12.8% 5.9% 12.6% JWH-250 CO 384 Neifeld et al.

7 RapidFire/MS-MS Screen 385 Figure 1. (a) The total ion chromatogram (TIC) for an entire plate run using the synthetic cannabinoid method. (b) The internal standard (JWH 073 N-CO d5) only for the plate shown in Part A. (c) JWH 018 N-CO only for the plate shown in Part A. The calibrator and controls are labeled, as are the two presumptive positive samples in the batch. All other samples are negative. (d) After the plate data file is "cut" into individual data files for each injection, the MRM transition for each compound in the sample can be observed. These peaks show the "cut" files for JWH 018 N-CO for individual samples in the plate. The arrows from the JWH 018 CO fragment to the MRM transitions correspond to the sample.

8 386 Neifeld et al. Table IX. Synthetic Cathinone LC Screen vs RF Screen Comparison for 18,000 Sample Study LC screen RF screen Number of compounds screened Sample amount 1 ml 50 µl Sample preparation method Liquid liquid extraction Dilute and shoot Instrument run time 9.5 min 9.1 s Total instrument time 2,850 h 55 h Total time for completion 3 months 4 weeks hydrolysis was assessed (β-glucuronidase from Escherichia coli) by using different amounts of the enzyme for varying incubation times and temperatures. Testing was performed using 50 and 100 µl of the sample with a range of 2,500 15,000 units of enzyme per milliliters of the sample. A JWH 018 CO conjugated control was used to determine the extent of hydrolysis. Room temperature incubation was examined for 20 min as well as incubation at 55 C for 10 and 20 min. However, the enzyme caused complete ion suppression and no analyte detection for the enzyme amounts and incubation times tested. There is not enough sample cleanup when using the RF method to remove the enzyme and allow detection of the analytes. Sodium hydroxide (Na) was evaluated as an enzyme alternative at varying concentrations, incubation times and temperatures. Hydrolysis was optimized when using 0.3 N Na with a room temperature incubation time of 15 min. A JWH 018 CO glucuronide control was analyzed using these conditions and resulted in at least 60% recovery of free analyte; this was considered acceptable for a screening assay. Using other Na concentrations and/or other incubation times and temperatures yielded less hydrolysis of the conjugated control. Different cartridges were evaluated to determine which packing material would work best for both classes of emerging drugs. Cartridges tested were C4, C8, C18 and cyano by analyzing a calibration curve as well as positive and negative controls. The C4 cartridge provided the best results overall for the majority of the compounds analyzed. The C4 cartridge is manufactured as A and A2 with A2 advertised to help reduce carryover between injections. Carryover was detected using the A cartridge and was reduced significantly using the A2 cartridge. For all compounds analyzed using the RF MS-MS methods for analysis of synthetic cathinones and synthetic cannabinoids, the precision was within 25% and the accuracy was within 30%. While some of the compounds are outside of the laboratory s 15% acceptable rate, this method is used for screening only. The higher precision and accuracy is, therefore, acceptable for this method. Troubleshooting As discussed earlier, sample preparation involved a dilution followed by RF MS-MS analysis. This sample preparation technique presented several issues that were resolved during method development. The particulates present in the urine caused some clogging issues with the RF system. When a clog occurred, the RF system did not always identify the error and would continue to sip samples without an error. This was due to the clog causing an interference with the sensor that indicated a sample had been sipped. When this happened, the sip sensor would have a sip time of <0.08 s, as opposed to the average sip time of 0.3 s when a sample was sipped accurately. The issue was identified because the sealant for each well had not been punctured, but there was a data file for each sample giving the appearance an injection occurred. The presence of sufficient sample and internal standard from the previous injection in the sample loop gave the appearance of a successful injection. To alleviate the issue, plates were centrifuged prior to being analyzed which helped prevent the clogs. Additionally, the sip height (the height that the sipper was off of the bottom of each well when sipping sample) was also increased to 3 mm. Extra washes, both aqueous and organic, were added at the end of each row (every 12 samples) to help remove any possible clogs that may be present. Agilent also added the capability of flushing the sample loop in the middle of a plate if a clog was detected in the system. Initially, it was thought the aluminum plate sealant being used was contributing to the clogs. The sipper was cut at an angle instead of straight to see if clogs were eliminated by using a sharper sipper tip to cut through the sealant. Instead of decreasing the number of clogs, this resulted in observed abundances that were about 10 times lower than those obtained with a straight cut. The sipper was re-cut straight, and the abundances increased. Varying sample volumes and dilution factors were explored to determine if using a more diluted urine sample would decrease clogging and matrix effect issues. To keep the automated sample preparation in the plate, it was determined that each well held a maximum of 300 µl. When a 1:10 dilution was used, abundances decreased and certain compounds could no longer be detected. When a 1:2 dilution was used, the RapidFire system became more frequently clogged and the source on the MS-MS had to be cleaned more often. An extra wash step in the RapidFire method (State #3) helped eliminate some of the clogging issues. However, this wash could only be added to the synthetic cannabinoid method. The extra wash for synthetic cathinones caused the majority of the compounds to be eluted from the cartridge prior to the valve switching to the MS, therefore, reducing sensitivity of the method. For synthetic cannabinoids, the extra wash step also helped reduce carryover between samples. RF MS-MS and LC MS-MS comparison The majority of the synthetic cathinone presumptive positive specimens that ultimately confirmed negative during the analysis of the 18,000 specimens produced screening results that indicated the presence of mephedrone/nec/buphedrone (153), α-ppp (173), 4-MEC (8), 5-APB (30) and combinations of these compounds (14). Alternate MRM transitions for these compounds were evaluated and used to analyze the presumptive positives for these compounds. The majority of the specimens were negative when using the alternate transitions (377 of the 378 samples). The alternate transitions were validated and replaced the primary transitions in the method. Similar issues were identified for JWH 073 CO during initial method validation. During method validation, 10 negative urines were analyzed and the primary transition used for JWH 073 CO ( ) resulted in a positive result from the matrix in most of the samples. Therefore, a second transition ( ) was used for JWH 073 CO. When analyzing previously confirmed positive samples using an IA kit, only samples containing JWH 073 CO and/or JWH 018 CO screened as presumptively positive. The kit was not designed

9 RapidFire/MS-MS Screen 387 to detect other structural classes of synthetic cannabinoids. When screening samples using RF MS-MS, 84 of 86 samples with concentrations above 2 ng/ml screened as presumptively positive for synthetic cannabinoids, resulting in a 97.6% positive rate. Using RF MS-MS instrumentation for the initial screen enabled the detection of 14 different synthetic cannabinoids in contrast to the IA screening assay, which identified only two compounds. Conclusion The RapidFire MS-MS system provides a quick, efficient screening method for the detection of synthetic cathinones and synthetic cannabinoids in urine. Sample preparation time is minimal and can be automated, which reduces the time and material required for each analysis. When using the RF MS-MS screening method, calculated cost per sample was approximately six times less than when using an LC MS-MS method. While the detection limits are higher compared with the LC MS-MS screen, the sensitivity is better than IA screening methods. Additionally, the RF MS-MS screen allows for identification of specific analytes in a sample as opposed to an analyte class when using IA. As newer synthetic cathinones and synthetic cannabinoids are identified in commercial products or biological specimens, they can be easily added to the existing RapidFire methods. The RF MS-MS technology presents a promising strategy for detecting these frequently modified NPSs in urine. Disclaimer The opinions or assertions presented hereafter are the private views of the authors and should not be construed as official or as reflecting the views of the Department of Defense, the Defense Health Agency or the Armed Forces Medical Examiner System. Funding This work was funded in part by ARP Sciences, LLC, Rockville, MD. References 1. Heltsley, R., Shelby, M.K., Crouch, D.J., Black, D.L., Robert, T.A., Marshall, L. et al. (2012) Prevalence of synthetic cannabinoids in U.S. athletes: initial findings. Journal of Analytical Toxicology, 36, Spiller, H.A., Ryan, M.L., Weston, R.G., Jansen, J. (2011) Clinical experience with and analytical confirmation of bath salts and legal highs (synthetic cathinones) in the United States. Clinical Toxicology (Philadelphia), 49, Pasin, D., Bidny, S., Fu, S. (2015) Analysis of new designer drugs in postmortem blood using high-resolution mass spectrometry. Journal of Analytical Toxicology, 39, Ammann, D., McLaren, J.M., Gerostamoulos, D., Beyer, J. (2012) Detection and quantitation of new designer drugs in human blood: part 2 designer cathinones. Journal of Analytical Toxicology, 36, Uralets,V.,Rana,S.,Morgan,S.,Ross,W.(2014)Testingfordesigner stimulants: metabolic profiles of 16 synthetic cathinones excreted free in human urine. Journal of Analytical Toxicology, 38, Regester, L.E., Chmiel, J.D., Holler, J.M., Vorce, S.P., Levine, B., Bosy, T.Z. (2015) Determination of designer drug cross-reactivity on five commercial immunoassay screening kits. Journal of Analytical Toxicology, 39, Petrie, M., Lynch, K.L., Ekins, S., Chang, J.S., Goetz, R.J., Wu, A.H.B. et al. (2013) Cross-reactivity studies and predictive modeling of bath salts and other amphetamine-type stimulants with amphetamine screening immunoassays. Clinical Toxicology, 51, Cody, J.T. (1990) Cross-reactivity of amphetamine analogues with Roche Abuscreen radioimmunoassay reagents. Journal of Analytical Toxicology, 14, Nieddu, M., Burrai, L., Trignano, C., Boatto, G. (2014) Cross-reactivities of 39 new amphetamine designer drugs on three abuse drugs urinary screening tests. Forensic Toxicology, 32, Holler, J.M., Vorce, S.P., Knittel, J.L., Malik-Wolf, B., Levine, B., Bosy, T.Z. (2014) Evaluation of designer amphetamine interference in GC MS amine confirmation procedures. Journal of Analytical Toxicology, 38, Prosser, J.M., Nelson, L.S. (2012) The toxicology of bath salts: a review of synthetic cathinones. Journal of Medicinal Toxicology, 8, Marinetti, L.J., Antonides, H.M. (2013) Analysis of synthetic cathinones commonly found in bath salts in human performance and postmortem toxicology: method development, drug distribution and interpretation of results. Journal of Analytical Toxicology, 37, Wyman, J.F., Lavins, E.S., Engelhart, D., Armstrong, E.J., Snell, K.D., Boggs, P.D. et al. (2013) Postmortem tissue distribution of MDPV following lethal intoxication by bath salts. Journal of Analytical Toxicology, 37, Shanks, K.G., Dahn, T., Behonick, G., Terrell, A. (2012) Analysis of first and second generation legal highs for synthetic cannabinoids and synthetic stimulants by ultra-performance liquid chromatography and time of flight mass spectrometry. Journal of Analytical Toxicology, 36, Mohr, A.L.A., Ofsa, B., Keil, A.M., Simon, J.R., McMullin, M., Logan, B. K. (2014) Enzyme-linked immunosorbent assay (ELISA) for the detection of use of the synthetic cannabinoid agonists UR-144 and XLR-11 in human urine. Journal of Analytical Toxicology, 38, Arntson, A., Ofsa, B., Lancaster, D., Simon, J.R., McMullin, M., Logan, B. (2013) Validation of a novel immunoassay for the detection of synthetic cannabinoids and metabolites in urine specimens. Journal of Analytical Toxicology, 37, Kronstrand, R., Roman, M., Andersson, M., Eklund, A. (2013) Toxicological findings of synthetic cannabinoids in recreational users. Journal of Analytical Toxicology, 37, Romm, M., Miller, V.P. Application Note: Ultrafast Screen for Synthetic Cannabinoids in Urine Using the Agilent RapidFire High-Throughput Mass Spectrometry System. Agilent Technologies, Parikh, N.R., Romm, M., Miller, V.P. Application Note: Ultrafast Screen for Bath Salts in Urine Using the Agilent RapidFire High-Throughput Mass Spectrometry System. Agilent Technologies, 2013.