Keywords

1 Introduction

Bioanalysis of new psychoactive substances (NPS) is very challenging due to the growing number of compounds available on the drugs of abuse market. In Europe, the number of NPS notified by the European Union (EU) Early Warning System coordinated by the European Monitoring Centre for Drugs and Drug Addiction (EMCDDA) has risen rapidly since 2008 (EMCDDA 2017). More than 700 NPS from different chemical and pharmacological classes are continuously monitored as part of the EMCDDA’s early-warning and toxicovigilance system. A significant proportion of detected substances are synthetic cathinones or synthetic cannabinoid receptor agonists (SCRAs), but synthetic opioids and benzodiazepines are on the rise since 2014. The first indication of a newly emerging compound on the market often arises from its identification in seized samples or from test purchases. Recommendations on such kind of analysis are published, for example, by the Scientific Working Group for the Analysis of Seized Drugs (SWGDRUG 2016). NPS initially circumvent existing legislation and are usually offered as legal alternative to traditional and thus controlled drugs of abuse. Consumers can easily obtain them, e.g., via Internet retailers, labeled, for instance, as research chemicals, plant food, or bath salts, often with the addition “not for human consumption.” The Internet provides a rich source of information for consumers as well as for health-care professionals or policy makers (Deluca et al. 2012). Even though NPS are initially not controlled and easily available, information about potentially harmful effects is frequently available (Logan et al. 2017; Nelson et al. 2014; Zamengo et al. 2014).

In the following, an overview is provided of recent developments in the field of bioanalytical methods of analysis applied to a variety of NPS including information on sample preparation, instrumentation, detection modes, data evaluation, and pitfalls. PubMed-listed and English-written original research papers and review articles published during 01 October 2012 and 30 September 2017 were considered including single- and multi-analyte procedures.

2 Methods

Literature research was performed in PubMed (National Center for Biotechnology Information, US National Library of Medicine, Bethesda, MD, USA) for English-written articles published online between 01 October 2012 and 30 September 2017. The search terms “new psychoactive substances” or “novel psychoactive substances” in combination with one of the key words “screening,” “detection,” “quantification,” “bioanal*,” “mass spectrom*,” or “quantification” in any field were used: ((new psychoactive substances OR novel psychoactive substances) AND (screening OR detection OR quantification OR bioanal* OR mass spectrom*)) AND (“2012/10/01” [PDAT]: “2017/09/30” [PDAT]). Both review articles and original research papers were considered but single case reports only if they were the only information source concerning the intake of a specific NPS and if an analytical confirmation was described.

3 General Aspects on Bioanalysis of NPS

Screening, identification, and quantification of NPS in different biosamples such as blood, urine, hair, oral fluid, and tissue samples are important in the clinical and forensic toxicology environment (Peters 2014). Biosampling strategies used for matrix alternatives to classic blood and urine were reviewed elsewhere (Mercolini and Protti 2016). Urine is the preferred matrix in various areas of analytical toxicology, such as abstinence monitoring, workplace drug testing, or where comprehensive screening purposes are required, as it can be obtained noninvasively in comparable large volumes and all substances and/or their metabolites are concentrated (Meyer and Maurer 2016; Peters 2014). Especially in point-of-care drug testing, immunoassay screening commonly provides a first indication of the presence of drugs (of abuse). However, only few immunoassays applicable to the detection of NPS were developed (Arntson et al. 2013; Barnes et al. 2015; Castaneto et al. 2015a; Mohr et al. 2014). A considerable challenge stems from the multiplicity of substances and chemical structures available but also from the time-consuming production of antibodies that takes often more time than the “half-life” of a specific NPS on the market (Favretto et al. 2013). More promising could be a structure-independent assay system recently reported for activity profiling of synthetic cannabinoids developed by Cannaert et al. (2016). They developed cell systems stably expressing the cannabinoid receptors CB1 or CB2 connected to a luciferase reporter system. The assay allowed for the measurement of receptor activation via β-arrestin involvement and bioluminescence, which was successfully applied to authentic urine samples obtained from users of SCRAs (Cannaert et al. 2017). This approach is independent of the compound structure but requires excretion of active substances into urine. Furthermore, phytocannabinoids are also detected, and further identification and confirmation steps are needed, preferably by mass spectrometry (MS). In immunoassays designed for detection of traditional drugs, some NPS might give positive results due to cross-reactivity, but that is dependent on the NPS class and concentration, which means that they are not identified reliably (Beck et al. 2014; Nieddu et al. 2016; O’Connor et al. 2016; Pettersson Bergstrand et al. 2017a; Swortwood et al. 2014).

Recent literature is consistent in the recommendation to use MS-based methods for NPS screening (Al-Saffar et al. 2013; Favretto et al. 2013; Franz et al. 2017a; Sundstrom et al. 2015). Only MS, usually deployed after implementation of a variety of separation methods, provides the high level of flexibility, sensitivity, and selectivity that is needed for robust and reliable detection of NPS (Meyer and Maurer 2016). However, lipophilic NPS are extensively metabolized, and thus, the analytical strategy has to consider metabolites as targets particularly in urine. According to several national guidelines, e.g., in Sweden, metabolite detection needs confirmation using reference substances, and the use of mass spectral reference libraries also containing the metabolites is highly recommended (Helfer et al. 2015; Wissenbach et al. 2011). In contrast to licensed therapeutic drugs, pharmacokinetic data for NPS are not available. Thus, metabolism studies are mandatory for developing screening methods. Metabolite data can be generated by using in vitro or in vivo models, or metabolites can be identified in samples obtained from authentic cases (Maurer and Meyer 2016; Richter et al. 2017; Schaefer et al. 2016; Welter-Luedeke and Maurer 2016). Further information can be found in the review and the corresponding chapter of this handbook by Meyer summarizing toxicodynamics and toxicokinetics of NPS (Meyer 2016).

MS-based screening procedure can be divided into targeted and untargeted procedures (Meyer and Maurer 2016). Targeted screenings, often combined with quantification, focus on a predefined set of analytes and are traditionally performed with low-resolution tandem mass spectrometry (MS/MS) devices using selected reaction monitoring (SRM) mode providing high selectivity and sensitivity. Covering only the targets, such procedures cannot detect unexpected or unknown compounds. In contrast, untargeted full scan screenings cover all analytes contained in the used library and allow retrospective data mining. They can also give indications of new compounds if they can be extracted, separated, and ionized. High-resolution (HR) MS/MS data often help to get an idea about the structure, particularly if known partial structures are indicated. Of course, the identity must be confirmed by reference standards. HRMS can also be used for data-independent acquisition (DIA), where all product ions are recorded regardless of the precursor ion, and data-dependent acquisition (DDA) mode, where preset criteria define the precursor ions for MS/MS spectral recording. Sundstrom et al. compared post-targeted DIA and pre-targeted DDA for urine drug screening based on quadrupole time-of-flight (QTOF) MS and discovered that DIA was more straightforward and the method was easier to deploy in casework and that the DDA approach with substance-specific collision energies produced informative product ion spectra suitable for occasional confirmatory analyses (Sundstrom et al. 2017).

4 Bioanalysis of NPS of Different Classes

Some methods for simultaneous detection of a broad range of NPS of different classes were published, and details are given in Table 1. Mainly stimulants but also SCRAs and hallucinogens were included in such targeted multi-analyte approaches. The blood screening procedure described by Adamovicz and Tokarczyk covered 143 NPS with a rapid and simple sample preparation and limits of detection (LOD) estimated for 104 compounds in the range 0.01–3.09 ng/mL (Adamowicz and Tokarczyk 2016). Vaiano et al. included 64 NPS in their method for blood prepared by protein precipitation (Vaiano et al. 2016). Lehmann et al. used automated solid-phase extraction (SPE) for detection of 69 NPS in serum, and total cycle time for one sample was only 11 min due to the interlacing between sample preparation and analysis (Lehmann et al. 2017). Odoardi et al. applied dispersive liquid-liquid microextraction using minimal amount of organic solvent as rapid, cheap, and efficient alternative sample preparation for blood (Odoardi et al. 2015). Ambach et al. successfully used dried blood spots as an alternative sampling strategy to screen for 64 NPS and also reported about substance stability in dried blood spots (Ambach et al. 2014). Three studies focusing on targeted screening in urine were identified, but unfortunately, only parent compounds and no metabolites were included, which is insufficient for comprehensive urine screening procedures as discussed earlier (Al-Saffar et al. 2013; Lee et al. 2016; Tang et al. 2014). Boumba et al. described a procedure for analysis of 132 NPS in hair, which provided LOD from 0.001 to 0.1 ng/mg hair (Boumba et al. 2017). LOD were comparable to them described in other studies (Montesano et al. 2017; Strano-Rossi et al. 2014).

Table 1 Biosamples, experimental setups, and highlights of multi-analyte approaches for detection of NPS of different classes
Full size table

The majority of these multi-analyte approaches were targeted screenings based on liquid chromatography (LC) coupled to MS/MS operating in SRM mode with triple quadrupole or quadrupole ion trap hybrid mass analyzers. Only three screening procedures based on HRMS were described (Montesano et al. 2016, 2017; Stephanson et al. 2017). Stephanson et al. performed first compound detection by extracted ion chromatograms from full scan and second confirmation by product reaction monitoring. Only in case of interferences in full scan, product reaction monitoring was already used as first step for some NPS (Stephanson et al. 2017). The studies published by Montesano et al. used both, targeted selected-ion monitoring and full scan (Montesano et al. 2016, 2017). The application of full scan also allowed for retrospective data mining.

5 Bioanalysis of NPS Stimulants

NPS acting as psychostimulants are derivatives of phenethylamine, amphetamine, or cathinone, and they belong to the most abundant representatives on the NPS market in the EU (Tyrkko et al. 2016). Some reviews summarized the existing knowledge about stimulants including analytical methods and detectability. Whereas Welter-Luedeke and Maurer focused on amphetamine derivatives with modified ring systems, Ellefsen et al. were interested in synthetic cathinones (Ellefsen et al. 2016; Welter-Luedeke and Maurer 2016). Alvarez et al. published a review about hair analysis of synthetic cathinones (Alvarez et al. 2017). Details of original research papers focusing on bioanalysis of stimulants can be found in Table 2. As part of the Swedish STRIDA project, five studies summarizing serum and urine concentrations of 15 different stimulants were published (Backberg et al. 2015c, 2016; Beck et al. 2015, 2017). The STRIDA project was initiated to monitor the occurrence and health hazards associated with NPS consumption in Sweden. Information about clinical effects originated from medical record data in laboratory confirmed cases of acute drug intoxications presenting at emergency departments or intensive care units all over Sweden and were collected to extend the knowledge about NPS, drug trends, and health risks over a longer time period (Helander et al. 2014b). Adamowicz et al. interpreted 3-MMC blood concentrations in forensic context based on analysis of 95 cases (Adamowicz et al. 2016). Maas et al. considered the identification of positional isomers to be an important issue in forensic toxicology due to differences in legal status or toxicity and developed a method for the separation of ortho-, meta-, and para-isomers of methylmethcathinone and methylethcathinone in serum samples (Maas et al. 2017). Grapp et al. reported MDPV serum concentrations determined in 23 cases (Grapp et al. 2017). Olesti et al. developed a quantification method for mephedrone in plasma and urine and applied it to samples taken from six volunteers following oral intake of 150 mg mephedrone as part of a randomized, double-blind, crossover controlled clinical trial (Olesti et al. 2017). The metabolic fate of two α-PVP derivatives was elucidated using human hepatocyte incubations and authentic urine specimens (Swortwood et al. 2016a, b). Concheiro et al. developed a procedure for simultaneous determination of 40 stimulants in urine (Concheiro et al. 2015). Unchanged parent compounds were chosen as targets, and only four metabolites were included, which might prove insufficient for urine screening, despite implementation of SPE for concentration during sample preparation. Two studies focused on detection of stimulants in hair samples (Lendoiro et al. 2017; Salomone et al. 2016). Ares et al. presented a screening procedure for oral fluid based on microextraction by packed sorbent (MEPS) followed by LC-MS/MS. MEPS, a miniaturized version of SPE, provided some advantages including short extraction times, reduced sample and solvent consumption, and facile operation and was shown to be a suitable sample preparation method for oral fluid (Ares et al. 2017). Williams et al. also developed and validated a method for detection of synthetic cathinones in oral fluid and prepared samples by simple precipitation (Williams et al. 2017). Furthermore, various stimulants were included in multi-analyte approaches mentioned above (Adamowicz and Tokarczyk 2016; Al-Saffar et al. 2013; Ambach et al. 2014; Boumba et al. 2017; Lee et al. 2016; Lehmann et al. 2017; Montesano et al. 2016, 2017; Odoardi et al. 2015; Stephanson et al. 2017; Strano-Rossi et al. 2014; Tang et al. 2014; Vaiano et al. 2016).

Table 2 Biosamples, experimental setups, and highlights of studies containing analytical methods for detection of stimulants
Full size table

6 Bioanalysis of Synthetic Cannabinoid Receptor Agonists

Together with stimulants, SCRAs form the largest group of compounds detected on the NPS market. Smokable, herbal products containing SCRAs, commonly referred to as “Spice,” are believed to share some biological effects of naturally occurring phytocannabinoids (Karila et al. 2016). However, numerous adverse events and fatalities linked to their use were reported most likely caused by distinct pharmacological properties enhancing their toxic profile. On the one hand, SCRAs were described as full cannabinoid receptor CB1 and CB2 agonists, in contrast to partial agonist properties of the primary psychoactive compound of marijuana Δ9-tetrahydrocannabinol (THC), and on the other hand, several metabolites have been reported to retain, at least in some cases, affinity and activity at cannabinoid receptors (Fantegrossi et al. 2014). Some confusion can be caused by nomenclature of SCRAs. For example, AMB-FUBINACA, MMB-FUBINACA, and FUB-AMB are different names of the same compound (Adams et al. 2017). Structurally different from THC, neither parent compound nor SCRA metabolites are usually detected with standard cannabinoid immunoassays. Even if development and validation of specialized immunoassays are possible, this is extremely time-consuming, and due to the structural differences within the group of SCRAs, only few compounds are detected, which are already replaced by others once the immunoassay is marketable (Castaneto et al. 2015b). However, immunoassays and the already discussed bioassay for activity profiling developed by Cannaert et al. must be complemented by MS-based confirmation analyses (Cannaert et al. 2016; Castaneto et al. 2015b). Castaneto et al. published a review about SCRA pharmacokinetics and detection methods in biological matrices. They mentioned that interpretation of MS-based results should include discussion of common metabolites formed by different SCRAs. Details of original research articles focusing on bioanalysis of SCRAs can be found in Table 3. Tynon et al. developed a screening and confirmation procedure for 34 SCRAs in whole blood (Tynon et al. 2017). Karinen et al. reported SCRA concentrations measured in blood samples of drivers suspected of impaired driving in Norway (Karinen et al. 2015). Protti et al. successfully used dried hematic microsamples for detection of SCRAs (Protti et al. 2017). Adams et al. identified AMB-FUBINACA, contained in the herbal incense “AK-47 24 Karat Gold,” as trigger of a mass intoxication in New York City by analyzing the incense product and biosamples of eight intoxicated individuals. However, the parent compound being a methyl ester could only be found in the incense product itself while only the de-esterified acid metabolite was detected in biosamples (Adams et al. 2017). Backberg et al. published a series of nine analytically confirmed intoxication with MDMB-CHMICA from the Swedish STRIDA project (Backberg et al. 2017). Franz et al. detected metabolites of the same SCRA in 818 authentic urine samples and analyzed furthermore in vitro incubations with human liver microsomes and smoke condensates (Franz et al. 2017b). For most SCRAs, the parent compounds are rarely detected in urine. Vikingsson et al. used authentic urine samples and incubations with human liver microsomes to identify urinary AB-FUBINACA metabolites suitable as targets for drug testing (Vikingsson et al. 2016). Mogler et al. described similar experiments for detecting 5F-MDMB-PICA metabolites (Mogler et al. 2017). Borg et al. developed an analytical method to detect metabolites of 32 SCRAs in urine specimens (Borg et al. 2017). This was the only multi-analyte approach used for urine screening that exclusively focused on metabolite detection. Salomone et al. published a screening procedure for detection of 23 SCRAs in hair covering only the parent compounds (Salomone et al. 2014). However, detecting only parent compounds complicates the differentiation of consumption from passive contamination (Castaneto et al. 2015b). Furthermore, some SCRAs or their metabolites were included in the multi-analyte approaches for NPS of different classes as described before (Adamowicz and Tokarczyk 2016; Boumba et al. 2017; Lee et al. 2016; Montesano et al. 2016, 2017; Odoardi et al. 2015; Strano-Rossi et al. 2014; Tang et al. 2014; Vaiano et al. 2016).

Table 3 Biosamples, experimental setups, and highlights of studies containing analytical methods for detection of synthetic cannabinoid receptor agonists
Full size table

7 Bioanalysis of Synthetic NPS Opioids

In both Europe and North America, the recent emergence of new synthetic opioids, mostly fentanyl derivatives, is causing considerable concern. Due to their high potency, they present serious health risks, not only to those who use them but also to those involved in their handling. Consequently, reports about nonfatal intoxications but also deaths caused by synthetic opioids are increasingly reported since 2012 (EMCDDA 2017). Due to structural differences to morphine, a detection of synthetic opioids with standard immunoassays designed for opiates is not possible (Helander et al. 2014a). Details of studies focusing on bioanalysis of synthetic opioids can be found in Table 4. Within the framework of the STRIDA project, three studies were published that described a total of 29 intoxication cases involving 10 different fentanyl derivatives (Backberg et al. 2015b; Helander et al. 2016, 2017). The same authors also published a report about a case series of nonfatal intoxications with MT-45 (Helander et al. 2014a). Papsun et al. developed and validated a method for MT-45 quantification in whole blood and Fleming et al. for detection and quantification of U-47700 including four authentic urine samples (Fleming et al. 2017; Papsun et al. 2016). Metabolites of fentanyl derivatives as targets for urine drug testing were identified by Steuer et al. (2017) and Watanabe et al. (2017). Noble et al. developed a targeted screening method to detect 50 fentanyl analogs in whole blood using LC-QTOF-MS and offered a validation for 13 compounds (Noble et al. 2017a).

Table 4 Biosamples, experimental setups, and highlights of studies containing analytical methods for detection of synthetic opioids
Full size table

8 Bioanalysis of Designer Benzodiazepines

Over the last few years, an increasing number of benzodiazepine NPS appeared. These so-called designer benzodiazepines are structurally related to clinically used benzodiazepines (Manchester et al. 2017). Owing to this structure similarity, cross-reactivity with immunoassay antibodies was observed, and some designer benzodiazepines could be detected in standard immunoassay drug screenings of both blood and urine samples (O’Connor et al. 2016; Pettersson Bergstrand et al. 2017a). However, false negative immunoassay screening results have also been encountered (Huppertz et al. 2017; Moosmann et al. 2013, 2014). Negative immunoassay results underline the need for MS-based analysis, but analytical results have to be interpreted with care. Licensed benzodiazepines can be identical to designer benzodiazepine metabolites (e.g., clonazepam detection after intake of cloniprazepam), or the designer benzodiazepine itself can have the same structure as metabolites of prescribed benzodiazepines (e.g., fonazepam, identical to norflunitrazepam) (Moosmann et al. 2016). Isomers such as diclazepam and 4-chlorodiazepam can also lead to misidentification. Furthermore, designer benzodiazepines are possibly used as therapeutics in some countries, such as phenazepam, which was a prescription drug in the former Soviet states, and etizolam, which was originated in Japan (Manchester et al. 2017).

The review article by Manchester et al. about the emergence of designer benzodiazepines contained also an overview of analytical methods employed for their identification in biological matrices (Manchester et al. 2017). Experimental setups and highlights of original research articles containing analytical methods for detection of designer benzodiazepines are summarized in Table 5. A combined targeted and nontargeted drug screening in whole blood by HRMS with data-independent acquisition was described by Mollerup et al. (2017). First, a targeted screening was performed with identification based on reference standard data. However, the authors stated that due to the continuous appearance of NPS, it is almost impossible to generate all reference standard data and to keep the analytical method updated permanently. To overcome this shortcoming, a subsequent nontargeted screening extracted information regarding previously unidentified peaks for additional drug identifications and metabolite confirmation. To test its applicability, blood was spiked with 11 low-dosed designer benzodiazepines, and 9 were tentatively identified in the nontargeted screening approach in the concentration range 0.005–0.1 mg/kg. Hoiseth et al. published blood concentrations of five designer benzodiazepines detected in 77 authentic samples of drugged drivers or other criminal offenders (Hoiseth et al. 2016). After a first screening for a broad repertoire of drugs of abuse and medicines, a second confirmation analysis with quantification was performed. The authors stated that reported concentrations could be helpful in future interpretation and assessment of designer benzodiazepine blood concentrations in forensic toxicology.

Table 5 Biosamples, experimental setups, and highlights of studies containing analytical methods for detection of designer benzodiazepines
Full size table

As previously discussed, metabolites have to be considered in urine screening (Manchester et al. 2017). Several studies describing detection of designer benzodiazepines in urine provided also information about formed metabolites (Huppertz et al. 2017; Meyer et al. 2016; Moosmann et al. 2013, 2014; Noble et al. 2017b; Pettersson Bergstrand et al. 2017b; Vikingsson et al. 2017). The studies performed by Moosmann et al. in 2013 and 2014 and Huppertz et al. in 2017 were very similar in their experimental setup but differed in the investigated compound. Each of them contained a self-experiment by one of the authors, and serum concentrations and metabolites in urine were described as well as elimination half-life and results of immunoassay screening (Huppertz et al. 2017; Moosmann et al. 2013, 2014). Noble et al. investigated the metabolic fate of flubromazolam by means of in vitro incubations with pooled human liver microsomes or recombinant cytochrome P450 isoforms. Metabolites were confirmed by analyses of authentic samples obtained from two forensic cases (Noble et al. 2017b). Vikingsson et al. used mice urine and human hepatocytes in addition to microsomes and authentic human urine samples to identify meclonazepam metabolites. Similar to therapeutic nitro-containing benzodiazepines, amino-meclonazepam and acetamino-meclonazepam were found to be the main metabolites in human urine, which were also found in mice urine and human primary hepatocytes. However, human liver microsomes were only capable of producing minor amounts of the amino metabolite (Vikingsson et al. 2017). Meyer et al. were also interested in the identification of main urinary metabolites of three nitrobenzodiazepines in authentic urine samples and compared results obtained by nanoLC-HRMS/MS to conventional ultra-high performance LC-HRMS/MS (Meyer et al. 2016). The nanoLC system was found to provide a higher abundance of all signals and consequently higher sensitivity allowing for detection of additional compounds. Whereas clonazolam and meclonazepam were mainly excreted as their amino and acetamino metabolites, nifoxipam was additionally detected as glucuronide. As the parent compounds were generally present in low concentrations or not found at all, the authors emphasized the need to involve metabolites in the used screening procedure and recommended 7-aminoclonazolam, 7-acetaminomeclonazepam, and 7-acetaminonifoxipam as suitable targets for urine drug testing. Concerning flubromazolam and pyrazolam, recommended targets for urine screening were described by Pettersson Bergstrand et al. (2017b). The same authors published a multicomponent LC-MS/MS method for determination and quantification of 11 designer benzodiazepines in urine after conjugate cleavage. The method was applied to analysis of 390 samples with a positive immunoassay benzodiazepine screening but a negative MS confirmation focusing on prescribed benzodiazepines. They could detect designer benzodiazepines in 40% of these cases (Pettersson Bergstrand et al. 2016).

9 Bioanalysis of Hallucinogenic NPS

Hallucinogens are a diverse group of drugs that alter perception, thoughts, and feelings. Naturally occurring compounds, such as tryptamines, produced by plants, mushrooms, or animals, as well as human-made hallucinogens such as lysergic acid diethylamide (LSD), are used (NIDA 2016). Synthetic tryptamines and LSD derivatives are available on the drugs of abuse market. Arylcyclohexamines, which include derivatives of ketamine or phencyclidine, and 2-methoxybenzyl-substituted amines, the so-called NBOMes (Kyriakou et al. 2015; Zawilska 2014), are also included in this particular group of compounds. Chemically, NBOMes are derivatives of phenethylamine, but examples of amphetamine- and tryptamine-based NBOMes have also been characterized (Brandt et al. 2015; Caspar et al. 2017c).

The review article published by Kyriakou et al. also contained analytical methods for the detection of NBOMes (Kyriakou et al. 2015). Experimental setups and highlights of original research articles containing analytical methods for detection of hallucinogens are summarized in Table 6. Since phenethylamine and lysergamide-based NPS are rather low-dosed, their reliable detection and identification in biosamples may be difficult. Caspar et al. published the simultaneous identification and quantification of low-dosed hallucinogens in blood plasma based on LC-HRMS including three LSD derivatives and five NBOMes (Caspar et al. 2018). Poklis et al. published a detection method for nine different NBOMes in urine specimens, applied it to authentic samples, and identified four different NBOMes (Poklis et al. 2014). Unfortunately, no metabolites were included, even though extensive information on metabolism of NBOMes have been described (Caspar et al. 2015, 2017; Temporal et al. 2017; Wohlfarth et al. 2017). Gee et al. described an LC-MS/MS method for determination of plasma and urine concentrations of 25B-NBOMe in a series of intoxication cases, although the use of 25D-NBOMe as an internal standard could be problematic in cases of polydrug abuse (Gee et al. 2016).

Table 6 Biosamples, experimental setups, and highlights of studies containing analytical methods for detection of hallucinogens
Full size table

The use of arylcyclohexamines among NPS users was also confirmed by two case series. Menzies et al. investigated the metabolic fate of the ketamine analog methoxetamine using in vitro incubations and authentic urine samples (Menzies et al. 2014). Backberg et al. reported 67 intoxication cases with phencyclidine derivatives that occurred in Sweden between July 2013 and March 2015 (Backberg et al. 2015a). For synthetic tryptamines, only few studies were published, focused mainly on metabolism (Caspar et al. 2017b; Grafinger et al. 2017; Michely et al. 2015, 2017). However, a case series from the STRIDA project describing intoxications with 5-(2-aminopropyl)indole (5-IT), a positional isomer of alpha-methyltryptamine (AMT, 3-IT), was published (Backberg et al. 2014). Similar to AMT, 5-IT was shown to be a reversible inhibitor of monoamine oxidase A and is difficult to be analytically differentiated from AMT (Elliott et al. 2013; Shulgin and Shulgin 1997; Wagmann et al. 2017). For this reason, 5-IT is mentioned here although being a stimulant (Shulgin and Shulgin 1997). Furthermore, hallucinogens, mainly tryptamines and NBOMes, but also some arylcyclohexamines, were included in the multi-analyte approaches discussed before (Adamowicz and Tokarczyk 2016; Al-Saffar et al. 2013; Ambach et al. 2014; Boumba et al. 2017; Lee et al. 2016; Lehmann et al. 2017; Montesano et al. 2017; Stephanson et al. 2017; Tang et al. 2014; Vaiano et al. 2016).

10 Conclusions

The ever-changing pool of NPS flooding the drugs of abuse market is certainly challenging for clinical and forensic toxicology. Nevertheless, the high number of research articles describing bioanalytical methods for detection of NPS published during the last 5 years demonstrated that this problem is widely known these days. While immunoassays were found to be rather inappropriate for detection of NPS, MS-based procedures proved to be suitable due to high flexibility, sensitivity, and selectivity. Especially HR devices are promising tools for identification of unknown compounds or very low-dosed substances. However, in comparison to low-resolution devices, these instruments are significantly more expensive, and data handling is more sophisticated. Nevertheless, their application will further increase during the next years if these disadvantages will be overcome.