How preclinical studies have influenced novel psychoactive substance legislation in the UK and Europe

Abstract

Novel psychoactive substances (NPS) are new drugs of abuse. Over the last 10 years 50–100 new NPS have been detected for the first time each year. This has led to numerous deaths and challenges to healthcare providers and law‐makers worldwide. We review preclinical studies of NPS and discuss how these studies have influenced legislative decisions. We focus on the UK legal system but include experiences from Europe. We reviewed manuscripts from 2008–2019 and have summarised the in vitro and in vivo data on NPS, highlighting how these studies define pharmacological mechanisms and how they might predict harm in humans. We found that only a small percentage of NPS have been examined in preclinical studies. Most preclinical studies of NPS focus on basic pharmacological mechanisms (46% of studies reviewed) and/or addictive liability (32%) rather than toxicity and harm (24%). Very few preclinical studies into NPS include data from chronic dosing schedules (9%) or female rodents (4%). We conclude that preclinical studies can predict harm to humans, but most of the predictions are based on basic pharmacology rather than demonstrated toxicity. Some of these studies have been used to make changes to the law in the UK and Europe and perhaps, because of the paucity of toxicology data, most NPS have been placed in the highly dangerous schedule 1 or Class A category in the UK. We suggest that in silico studies and high throughput toxicology screens might be the most economical way forward to rapidly screen the health harms of NPS.

1. INTRODUCTION

Over the last decade, the drug scene has completely changed. In 2018 a new novel psychoactive substance (NPS) was reported to the European Commission Early Warning System every week, with 55 NPS reported in total. This is a much lower number than that reported in 2013–2015.1 Despite the fact that the observed decline could have resulted from many factors such as international legislation or possible lack of innovation, many substances continued to resurge globally,1 indicating a new established illicit repertoire. Other approaches employed to capture the true number of NPS that are globally available includes using a 24/7 web crawler such as the NPS.Finder.2 This novel approach showed that the online scenario includes over 4000 unique psychoactive molecules of interest to psychonauts, a number that is approximately 4‐fold the number being reported to the known NPS databases (European Monitoring Centre on Drugs and Drug Addiction [EMCDDA] and United Nations Office on Drugs and Crimes).2

In the UK, NPS were originally known as “psychoactive drugs which are not prohibited by the United Nations Single Convention on Narcotic Drugs or by the Misuse of Drugs Act 1971, and which people in the UK are seeking for intoxicant use”.3 They were therefore initially emerging to circumvent international legislation. In the UK, the Psychoactive Substances Act 2016 broadly defined them as “any substance, which is capable of producing a psychoactive effect”,4 covering any substance that is not controlled by the Misuse of Drugs Act 1971. Under the latter definition, the UK Home Office issued a forensic strategy to test for the psychoactivity of an unknown substance,5 which is key for prosecution in the UK.6 In this respect, in vitro testing was recommended by the Advisory Council on the Misuse of Drugs (ACMD) to prove if a substance is psychoactive. This mandates positive receptor and functional assays at 1 or more of the following receptors or transporters: https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=56 (cannabinoid receptor type 1), https://www.guidetopharmacology.org/GRAC/FamilyDisplayForward?familyId=72 (γ‐aminobutyric acid A receptor), https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=6 (serotonin 2A receptor)_, https://www.guidetopharmacology.org/GRAC/FamilyDisplayForward?familyId=75 (N‐methyl‐D‐aspartate receptor), https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=319 receptor and https://www.guidetopharmacology.org/GRAC/FamilyDisplayForward?familyId=176 (https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=940, https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=5, https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=484). However, since in vitro testing may not be suitable for all substances (e.g. nitrous oxide and solvents), in vivo testing has been recommended as an alternative. in vitro testing can predict human doses for stimulants and psychedelics, for example7 but cannot easily predict human pharmacokinetics or brain penetration. Information may also be retrieved from scientific literature if in vitro/in vivo testing cannot be performed.5

Despite the apparent decline in the numbers of emerging NPS, research is lagging behind the rapid changes in this complex market. Case reports including clinical and forensic cases, and emergency and mental health admissions have shown that NPS are associated with unpredictable acute adverse events, serious harms and even deaths.8 Knowledge of long‐term effects following chronic use of NPS is lacking and systems that capture harms resulting from NPS are still patchy.9

Animal models have been previously validated and have played pivotal roles in preclinical research informing the neurobiological, psychopathological, behavioural and aetiological aspects underlying drug addiction and acute/chronic drug use.10, 11 They have long been used to inform on the various phases of drug addiction including drug self‐administration patterns, behavioural criteria that define vulnerability to drug abuse, conditioned place preference, craving and relapse, escalation of drug‐taking, impulsivity, continued drug‐taking despite adverse drug reactions, physical dependence and other neuropsychopathological aspects, which lead from a voluntary drug‐taking to compulsive behaviours.10, 11, 12, 13, 14 Furthermore, preclinical studies have contributed to understanding the toxicity spectrum associated with some traditional drugs of abuse including knowledge of neurotoxic effects seen with https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=4804 15 and https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=4574,16 as well as non‐neurological toxic effects such as MDMA‐induced heart‐valve issues17 and bladder problems associated with https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=4233 use.18

Recent reviews of NPS included studies which focussed on toxicity such as Assi et al.19 who reviewed the literature between 2007–2015 for NPS toxicity, yielding 20 studies and concluded that the harmful effects of NPS could be severe or lethal. Tracy et al.20 summarised NPS‐induced health harms as deduced from available clinical data, while Evans‐Brown and Sedefov reviewed the NPS risk assessment protocols across the EU.21

Here we review the preclinical studies of NPS over the last 10 years and summarise the findings from the main classes of NPS. We then examine the evidence that has been used in changing the legal status of NPS focussing on the UK with some examples from Europe.

2. MATERIALS AND METHODS

A literature search was performed on PubMed and Web‐of‐Science (Medline), from January 2008 to November 2019. We employed database‐specific search strategies with multiple keywords utilising word truncation/wild card symbols and index terms as appropriate for each database. The literature search was not undertaken using individually named NPS due to their large number, up to 4000 by some estimates2; consequently, not all pertinent studies may be identified.

Each identified article was categorised into the following topics (based on title and abstract review): pharmacological and behavioural experiments (preclinical studies), human studies (clinical), detection and identification (forensic), legal status (legal), epidemiology, comments, letters, replies (reports), and reviews.

2.1. Search strategy and study selection of NPS

The following search terms were used in PubMed:

(((((((legal high[Title/Abstract]) OR novel psychoactive substance[Title/Abstract]) OR new psychoactive substance[Title/Abstract]) OR bath salt[Title/Abstract]) OR designer drug[Title/Abstract]) OR plant food[Title/Abstract])) AND ((((((((((in‐vitro[Title/Abstract]) OR in‐vivo[Title/Abstract]) OR cell[Title/Abstract]) OR slice[Title/Abstract]) OR anesthetized[Title/Abstract]) OR anaesthetised[Title/Abstract]) OR rat[Title/Abstract]) OR mouse[Title/Abstract]) OR murine[Title/Abstract]) OR rodent[Title/Abstract]) NOT review.

In Web‐of‐Science, we selected the Medline database with the following search terms:

TI = (legal highs* or novel psychoactive substance* or new psychoactive substance* or bath salt* or designer drug* or plant food*) and MH = (in‐vivo* or in‐vitro* or cell* or slice* or anesthetised* or rat* or mouse* or rodents*).

The literature search yielded 335 articles from both PubMed and Web‐of‐Science (Figure 1). An additional 22 articles were deemed to be relevant. Forty‐seven articles were excluded due to duplicates. In total, 310 articles were screened for inclusion, and 66 were excluded as they were deemed irrelevant; 205 articles were screened for eligibility, and 136 were excluded in the final screening because they did not meet the following criteria: (i) NPS was not a major outcome; (ii) the article was a clinical study; (iii) the article was a forensic study; or (iv) the article was a review. In total, 108 journal articles from the original search were used in the final analysis (Figure 1).

Figure 1.

Flow diagram showing how articles were screened after PubMed and Medline searches. NPS = novel psychoactive substance

2.2. Nomenclature of targets and ligands

Key protein targets and ligands in this article are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY.

3. SUMMARY OF RESULTS FROM NPS SEARCHES

After excluding ineligible manuscripts, we found 55 animal (in vitro or in vivo) studies examining cathinones, 25 studies on stimulants, only 8 studies on synthetic cannabinoid receptor agonists (SCRAs) and 20 studies on other NPS. We have grouped the types of studies undertaken into 4 broad areas: (i) behavioural models of addiction; (ii) neurochemistry/neuropharmacology of addiction; (iii) toxicity; and (iv) cognitive dysfunction.

3.1. Cathinones

The most commonly studied cathinone was MDPV (3,4‐methylenedioxypyrovalerone; 25 studies [45%]), followed by mephedrone and methylone (16 studies [29%]) and α‐PVP (α‐pyrrolidinovalerophenone; 7 studies [13%]). All other cathinones studied were only examined in 3 or fewer manuscripts: (Table 1). Most studies examined these substances in models of addiction such as increased locomotion, locomotor sensitisation, conditioned place preference, drug self‐administration, increased intracranial brain stimulation and withdrawal effects; there were 31 (56%) such studies. The next most common type of study were neurochemical assays associated with dopamine, serotonin or noradrenaline; these mostly examined changes in these transmitter levels or effects on their transporters; there were 17 such studies (31%). The next most common type of study was related to toxicity: cell death, cell toxicity, reactive oxygen species or hyperthermia, and there were 11 such studies (20%). Three studies (5%) examined cardiovascular issues such as blood pressure or heart rate, 1 (2%) looked at hepatotoxicity, 3 (5%) looked at cognitive deficits such as memory impairment and 1 study (2%) examined the effects of cathinones on rodent offspring. Only 2 studies (4%) examined chronic effects of cathinones and only 2 studies (4%) reported using female rodents.

Table 1.

Pre‐clinical studies examining cathinone novel psychoactive substance (NPS) pharmacology. The most recent manuscripts are presented first

Study	DOI	NPS	Dose/conc	Species/sex	Age/weight	Samples	Assay	Main results
Dias da Silva, D. et al., 201922	10.1007/s00204‐019‐02539‐x	Metaphedrone	31 nM to 10mM 1 μM,10 μM, 100 μM and 500 μM	Male Wistar Han rats	150–250 g	Hepatocytes	Cytotoxic assays	↑ autophagic and apoptotic/necrotic mechanisms
Gannon, B.M. et al., 201923	10.1007/s00213‐018‐5046‐x	MDPV and methylone	0.32 mg/kg/inf	Male Sprague–Dawley rats	275–300 g		Self‐administration	Methylone + caffeine results in an enhanced reinforcing effectiveness compared to methylone alone
Luethi, D. et al., 201924	10.1177/0269881119844185	Methylone and MDPV				Human embryonic kidney 293 cells	Monoamine transporter assays	↑ potency to inhibit NA uptake with methylone compared with DA and 5‐HT uptake. ↓ transporter inhibition potencies with N‐demethylation of methylone. = NA and DA uptake inhibition potencies with O‐demethylenation of methylone and MDPV.
Blough, B.E. et al., 201925	10.1007/s00213‐018‐5063‐9	Methcathinone analogs		Male Sprague–Dawley rats	300–400 g	Synaptosomes		Compounds substituted at the 2‐position (ortho) were primarily dopaminergic Compounds Substituted at the 4‐position (Para) were found to be more serotonergic
Zhou, X. et al., 201926	10.3390/ijms20071561	3‐methylmethcathinone, mephedrone, methylone, MDPV, α‐PVP, and naphyrone				C2C12 myoblasts	Cytotoxic assays	All cathinones showed cell membrane integrity, depleted ATP levels, and increased mitochondrial superoxide concentrations α‐PVP and naphyrone impaired basal and maximal cellular respiration
Gerecsei, L.I. et al., 201927		MDPV		Mice		Neural tube	Cytotoxic assays, western blotting	↑ apoptotic cells in the pallium and in the subpallium ↓NR2B expression in the treated group
Kolesnikova, T.O. et al., 201928	10.1016/j.ntt.2019.02.001	α‐PVP	1, 5, 25 and 50 mg/L for 20 min 1, 5 and 10 mg/L for 7 d	Zebrafish			Drug exposure HPLC	α‐PVP psychostimulant effects at 5, 25 and 50 mg/L Hypolocomotion and repeated withdrawal after a 7‐day chronic treatment
Mayer, F.P. et al., 201929	10.1016/j.neuropharm.2018.12.032	Enantiomers of nor‐mephedrone, 4‐hydroxytolyl‐mephedrone (4‐OH‐mephedrone) and dihydro‐mephedrone				Human urine	Monoamine transporter assays	↑ potency at SERT with S‐enantiomers of nor‐mephedrone and 4‐OH‐mephedrone than the corresponding R‐enantiomers. ↓ effectivity as releaser at SERT with R‐enantiomers.
Eshleman, A.J. et al., 201930	10.1007/s00213‐018‐5059‐5	22 cathinones, 6 benzofurans and 1 stimulant				Human embryonic kidney 293 cells	Monoamine transporter assays	Most α‐pyrrolidinophenones had high hDAT potencies and selectivities 4‐cl‐ethcathinone and 3,4‐methylenedioxy‐N‐propylcathinone had higher hSERT selectivity Benzofurans had low hDAT selectivity and were releasers at hSERT
Yoon, H.S. et al., 201931	10.1016/j.neuint.2018.12.005	α‐PVT	20 mg/kg (i.p.)	Rats		Nucleus accumbens	Behavioural experiments, western blotting	↑ locomotor sensitization ↓ phosphorylation levels of GSK3β
Atehortua‐Martinez, L.A. et al., 201932	10.1177/0269881118822151	MDPV	3 mg/kg (i.p.)	Male Sprague–Dawley rats	8 wk 250–300 g	Nucleus accumbens, caudate putamen and prefrontal cortex	Behavioural experiments, western blotting	↑ stereotypies and open arm entries in the elevated plus maze with acute administration ↑ ΔFosB with chronic administration
Colon‐Perez, L.M. et al., 201833	10.1016/j.neuropharm.2018.04.031	MDPV	1 and 3 mg/kg (i.p.)	Male long Evans rats	250–300 g	Whole brain striatum	Functional magnetic resonance imaging immunoblotting DAT assay social interaction test ultrasonic vocalization intracranial self‐stimulation	↓ strength of correlated neural activity at 1 h ↓ striatal DAT at 24 h and caused a shift in subcellular levels and distribution of DAT, TH and VMAT2 ↓ ICSS thresholds at 1 h ↑ 50 kHz USV calls at 1 h ↓ social interactions
Oliver, C.F. et al., 201834	10.1016/j.drugalcdep.2018.01.013	MDPV	2 mg/kg (i.p.)	Male Sprague–Dawley rats	275–300 g		Locomotor activity, CPP, ultrasonic vocalization	AMD3100 reduced MDPV‐induced locomotor activation, conditioned place preference and modulated MDPV‐induced increase at 50 k‐Hz USV calls.
Luethi, D. et al., 201835	10.1016/j.neuropharm.2017.07.026	5‐IT, 4‐MA, 3‐MMC, N‐methyl‐2‐AI and MMAI				Human embryonic kidney 293 cells	Monoamine transporter assays	4‐MA, and MMAI entactogenic effects And 3‐MMC, 5‐IT, and N‐methyl‐2‐AI have stimulant‐type properties
Siedlecka‐Kroplewska, K. et al., 201836	10.1007/s12640‐018‐9898‐y	3‐Fluoromethcathinone	10 μM	Mice		Hippocampal HT22 cell line	Flow cytometry, western blotting	↑ intracellular production of reactive oxygen species ↑ formation of autophagic vacuoles ↓ level of p62/SQSTM1 protein
Zwartsen, A. et al., 201837	10.1016/j.neuro.2018.03.007	PMMA, α‐PVP, methylone and MDPV	PMMA = 10‐1000 μM α‐PVP = 1–300 μM Methylone = 1–1000 μM MDPV = 1–1000 μM	Male Wistar rats	Postnatal day 0–1	Cortical neurons culture	Multi‐well microelectrodes arrays	↓ the weighted mean firing rate and weighted mean burst rate
Gerecsei. Et al., 201838	10.3389/fnins.2018.00027	MDPV	10 mg/kg (s.c.)	C57Bl/6 J mice		Posterior intralaminar complex of the thalamus medial preoptic nucleus	Open field test, grip strength test, force plate actometry, pup retrieval test, nest building of pregnant mothers, in situ hybridization	↓ birth rate, survival of offspring and maternal care in the drug‐treated animals ↑ locomotor activity of the pups in the MDPV treated group
Gannon, B.M. et al., 201739	10.1097/FBP.0000000000000315	MDPV and α‐PVP	0.032 mg/kg/infusion (i.v.)	Male Sprague–Dawley rats	275–300 g		Self‐administration	Both enantiomers of MDPV and α‐PVP function as highly effective reinforcers.
Philogene‐Khalid, H.L. et al., 201740	10.1021/acschemneuro.7b00212	R‐MEPH and S‐MEPH	0.25, 0.50, 2.00 mg/kg/inf (i.v.)	Rats			Self‐administration ultrasonic vocalization	↑ break points rats trained to self‐administer R‐MEPH and greatest rates of 50 kHz ultrasonic vocalization than S‐MEPH
Nelson, K.H. et al., 201741	10.1016/j.pbb.2017.04.003	α‐PVP	0.3, 1 and 3 mg/kg (i.p.)	Male Sprague–Dawley rats	321–445 g		Conditioned place preference, conditioned taste avoidance	α‐PVP induced dose‐dependent taste avoidance as well as significant increases in time spent on the drug‐paired side (no dose‐dependent)
Gannon, B.M. et al., 201742	10.1016/j.drugalcdep.2017.06.031	MDPV	0.03, 0.10, 0.30, and 1.00 mg/mL	Male NIH Swiss mice	20–25 g		Two‐bottle choice procedures, radiotelemetry, conditioned place preference	↑MDPV preference solution when was paired with quinine but not with water ↑ escalate consumption with chronic (10 d) access at 0.30 mg/mL MDPV ↑ CPP at 0.30 mg/mL MDPV with a magnitude similar to the preference observed following intraperitoneal administration of MDPV ↑ locomotor activity at 0.1–1.0 mg/mL MDPV
Lantz, S.M. et al., 201743	10.1016/j.neulet.2017.06.059	Phthalimide	10–1000 μM	Rats		PC12 cells	Cytotoxicity assay	↑death cell after CP exposition CP alters mitochondrial function and DA and 5‐HT levels
Luethi, D. et al., 201744	10.1016/j.tox.2017.06.004	Bupropion, MDPV, mephedrone and naphyrone	0.5–2 mM			HepG2 cells HepaRG cells	Cytotoxicity assay	Bupropion, MDPV, mephedrone and naphyrone are mitochondrial toxicants
Grecco, G.G. et al., 201745	10.1016/j.taap.2017.05.010	Methylone, butylone and pentylone	20 mg/kg (s.c)	Male Sprague–Dawley rats		Blood samples CSF samples	Microdialysis HPLC	↑Cmax and AUC0‐∞, and the longest t1/2 in the plasma with pentylone ↑Cmax and AUC0‐∞ in the CNS with methylone and butylone
Elmore, J.S. et al., 201746	10.1038/npp.2016.213	Methylone	3, 6 and 12 mg/kg	Male Sprague–Dawley rats	250–300 g	Synaptosomes	Locomotor activity and temperature monoamine transporter assays in vivo microdialysis	↑ motor activation at 12 mg/kg methylone ↓ core temperature at 3 and 6 mg/kg but showed biphasic effects at 12 mg/kg Methylone acted as a fully efficacious substrate‐type releaser at DAT, NET and SERT ↑ brain extracellular dopamine and 5‐HT in vivo
McLaughlin, G. et al., 201747	10.1002/dta.2053	Mexedrone	Dose–response curve	Male Sprague–Dawley rats	250–300 g	Synaptosomes	Monoamine transporter assays	Mexedrone and N‐methoxymephedrone showed comparable potency at DAT but the latter compound was more potent at NET and SERT
Schindler, C.W. et al., 201648	10.1111/bph.13640	MDPV	0.3–3.0 mg/kg	Male Sprague–Dawley rats			Telemetry	S(+) enantiomer appeared to mediate the cardiovascular effects of MDPV
Berquist, M.D. et al., 201649	10.1016/j.drugalcdep.2016.05.001.	4‐MMC and MDPV	0.5 mg/kg (i.p.) 0.5, 1.0, or 2.0 mg/kg (i.p.)	Male Sprague–Dawley rats			Locomotor activity	Locomotor responses sensitize to MDPV and to certain mixtures of MDPV and 4‐MMC following repeated dosing
Schindler, C.W. et al., 201650	10.1007/s00213‐015‐4057‐0.	MDPV and methylone	0.03 mg/kg/inj 0.3 or 0.5 mg/kg/inj	Male Sprague–Dawley rats	300–400 g		Self‐administration microdialysis studies	MDPV self‐administration was acquired rapidly compare with methylone. MDPV (0.1 and 0.3 mg/kg) increased extracellular dopamine while i.v. methylone (1 and 3 mg/kg) increased extracellular dopamine and 5‐HT
Valente, M.J. et al., 201651	10.1093/toxsci/kfw105	Methylone, MDPV, pentedrone and 4‐MEC	0.05 to 10 or 20mM	Male Wistar rats	210–250 g	HepaRG	Cytotoxic assay	MDPV and pentedrone were the most cytotoxic. All cathinones triggered significant caspase activation and apoptosis
Colon‐Perez, L.M. et al., 201652	10.1038/npp.2016.40	MDPV	0.3, 1.0, or 3.0 mg/kg	Male long Evans rats	250–300 g	Prefrontal cortex	Functional magnetic resonance imaging	MDPV dose‐dependently reduced functional connectivity between frontal cortical and striatal areas Dopamine receptor blockade did not prevent the MDPV‐induced decrease in functional connectivity.
Gannon, B.M. et al., 201653	10.1124/jpet.115.229500	MDPV	0.01, 0.03, 0.10, 0.30, 1, 3, 10 and 30 mg/kg (i.p.)	Male NIH Swiss mice	20 – 25 g		Drug discrimination, radiotelemetry	S(+)‐MDPV was most potent to fully substitution for the cocaine training dose ↑locomotion after doses of S(+)‐MDPV and racemic MDPV
López‐Arnau, R. et al., 201554	10.1016/j.taap.2015.03.015	Mephedrone	25 mg/kg (s.c.)	Male Sprague–Dawley rats adolescent	115–130 g 5 wk	Hippocampus, striatum and frontal cortex	Radioligand binding experiments, western blotting, Morris water maze	↓ densities of dopamine and serotonin transporters without microgliosis ↓ expression of tyrosine hydroxylase and tryptophan hydroxylase 2. Impairment of the reference memory in the Morris water maze
Kiyatkin, E.A. et al., 201555	10.1038/npp.2014.191	MDPV and Methylone	1, 3, and 9 mg/kg (s.c.) 0.1, 0.3, and 1.0 mg/kg (s.c.)	Male long Evans rats	3–4 mo		Social interaction	Methylone and MDPV dose‐dependently increased brain temperature
Hutchinson, C.V. et al., 201556	10.1016/j.neulet.2015.03.021	Mephedrone	1 or 10 μM	Flatworm planaria			Conditioned place preference hypo‐locomotion	↑ preference to cocaine at both 1 and 10 μM ↓ locomotion after mephedrone withdrawal
Lopez‐Arnau, R. et al., 201457	10.1177/0269881114548439	Methylone	4 × 20 mg/kg (s.c.)	Male Sprague–Dawley rats	125–175 g 46 wk	Hippocampus, striatum and frontal cortex	Morris water maze, microglia activation	Hyperthermia, Memory loss
Adam, A. et al., 201458	10.1016/j.neuro.2014.07.004	MDPV	10 mg/kg	Male and female mice C57	7 d and 16 wk	Coronal brain slices	Caspase‐3	↑ neuronal apoptosis in young but not adult ↑ locomotion
Den Hollander, B. et al., 201459	10.1093/toxsci/kfu108	β‐Keto amphetamine and 4‐MMC		Male mice C57BL/6 J	8 wk	SH‐SY5Y neuroblastoma cells	Cytotoxicity assay	Cytotoxic
Opacka‐Juffry, J. et al., 201460	10.1016/j.pnpbp.2014.04.009	Mephedrone	0.3–30 μM	Male Wistar rats	8 wk	Nucleus accumbens	RTI‐121 binding, voltammetry	Displaced RTI‐121 and caused reverse transport of DA
Simmler, L.D. et al., 201461	10.1016/j.neuropharm.2013.11.008	Methedrone, 4‐MEC, 3‐FMC, pentylone, ethcathinone, buphedrone, pentedrone and N,N‐dimethylcathinone				Human embryonic kidney 293 cells	Monoamine transporter assay	All the cathinones were potent NA uptake inhibitors but different in DA vs 5‐HT None of the cathinones bound to rodent trace amine‐associated receptor 1
Kaizaki, 201462		α‐PVP	25 mg/kg oral	Male balbC mice	8 wk		Locomotion microdialysis	↑ release of DA and locomotion
Marusich, J.A. et al., 201463	10.1016/j.neuropharm.2014.02.016.	MDPV, α‐PVP, α‐PBP and α‐PPP	0.3–3.0 mg/kg MDPV, 1.0–10.0 mg/kg α‐PVP, 1.0–10.0 mg/kg α‐PBP, and 3.0–30.0 mg/kg α‐PPP	Male Sprague–Dawley rats	300–400 g	Synaptosomes	Transporter uptake and release assays locomotor activity	α‐PVP is a potent uptake blocker at dopamine and norepinephrine transporters α‐PBP and α‐PPP are also catecholamine transporter blockers but display reduced potency ↑ locomotor activity with all of them
Watterson, L.R. et al., 201464	10.1093/ijnp/pyu014	α‐PVP and 4‐MEC	1, 3, 10, 30, mg/kg 0.1, 0.3, 1, and 3 mg/kg	Male Sprague–Dawley rats	250 g		Intracranial self‐stimulation in rats	↑ intracranial self‐stimulation threshold reductions similar to that of methamphetamine
Bonano, J.S. et al., 201465	10.1007/s00213‐013‐3223‐5.	MDPV, methylone andmephedrone	0.32, 1 and 3.2 1, 3.2 and 10	Male Sprague–Dawley rats	314–387 g		Intracranial self‐stimulation	Methcathinone was the most potent compound, and MDPV was the longest acting compound
Gregg, R.A. et al., 201366	10.1097/FBP.0000000000000006.	Mephedrone	15 mg/kg (i.p.)	Male Sprague–Dawley rats	260–290 g		Locomotor activity	↑ cocaine‐induced locomotor activation by prior MEPH exposure
Gregg, R.A. et al., 201367	10.1016/j.drugalcdep.2013.06.014.	Mephedrone	15 mg/kg and 30 mg/kg	Male Sprague–Dawley rats	260–290 g		Sensitization paradigms	↑ repetitive movement by MEPH challenge compared to acute MEPH exposure in both paradigms
Gatch, M.B. et al., 201368	10.1097/FBP.0b013e328364166d.	MDPV, methylone, mephedrone, naphyrone, flephedrone and butylone	1, 3, 10 or 30 mg/kg 0.3, 1, 3, 10 or 30 mg/kg 3, 10, 30 or 100 mg/kg	Male Swiss–Webster mice	10 wk		Locomotor activity Discriminative stimulus effects	MDPV and naphyrone produced locomotor stimulant effects that lasted much longer than cocaine or methamphetamine all compounds fully substituted for the discriminative stimulus effects of cocaine and methamphetamine
Cameron, K. et al., 201369	10.1007/s00213‐013‐2967‐2.	MDPV andmephedrone		Female Xenopus laevis		Oocytes	Electrophysiological recordings	Mephedrone is a dopamine‐releasing agent and MDPV behaves as a cocaine‐like reuptake inhibitor of dopamine.
Baumann, M.H. et al., 201370	10.1038/npp.2012.204	MDPV	0.1–0.3 mg/kg, (i.v.) 0.1–3.0 mg/kg, (s.c.)	Male Sprague–Dawley rats male CB57/BL6 mice	300–400 g 25–35 g	Synaptosomes striatum nucleus accumbens	Uptake and release assays, fast‐scan cyclic voltammetry, in vivo microdialysis, telemetry, locomotor activity	↑ amplitude of the dopamine signal ↑ extracellular concentrations of dopamine in the nucleus accumbens ↑ locomotor activation, tachycardia, and hypertension
den Hollander, B. et al., 201371	10.1016/j.pbb.2012.10.006	Methylone or mephedrone	30 mg/kg, twice daily for 4 d	Male C57BL/J6 mice male Wistar rats	8 wk	Frontal cortex, striatum and hippocampus	Behavioural tests, neurotransmitter and transporters levels	↓ working memory with MEPH ↑ body temperature ↓ 5‐HT levels in the frontal cortex, striatum and hippocampus with methylone
Lopez‐Arnau, R. et al., 201372	10.1016/j.pnpbp.2013.04.007	Methylone	10–30 mg/kg (i.v. or oral)	Male Sprague–Dawley rats	225–250 g		Locomotion + pharmacokinetic analysis	↑locomotion
Cozzi, N.V. et al., 201373	10.1016/j.ejphar.2012.11.008	Methcathinone, 2‐TFMAP, 3‐TFMAP and 4‐TFMAP	Methcathinone: 0.3 mg/kg and 1.0 mg/kg 4‐TFMAP: 1.0 mg/kg and 3.0 mg/kg	Male Sprague–Dawley rats	300–350 g	Human platelets HEK293 cells C6 glioma cells Synaptosomes Accumbens	Monoamine transporter assay Microdialysis locomotor activity	↑ uptake inhibition and release with 3‐TFMAP and 4–TFMAP at SERT compared with methcathinone ↑ 5‐HT extracellular level with 4‐TFMAP but doesn't affect the locomotor activity
Lisek, R., 201274	10.1016/j.drugalcdep.2012.04.021.	Mephedrone	3, 5, 10, 30 mg/kg (i.p.) 30 mg/kg (i.p.)	Male Sprague–Dawley rats CD‐1 mice	225–275 g 25 – 30 g		Locomotor activity conditioned place preference	↑ ambulatory activity in rats and was inhibited by pretreatment with SCH 23390 and enhanced by pretreatment with sulpiride ↑ CPP by MEPH
Meng, H. et al., 201275	10.1016/j.toxlet.2011.10.010	Mephedrone	0.3–15 mg/kg	Chinese hamster Guinea pigs		Chinese hamster ovary cells single ventricular myocytes	Patch‐clamp myocyte action potential echocardio	↑HR and ↑BP
Baumann, M.H. et al., 2012 70	10.1038/npp.2011.304	Mephedrone and Methylone	0.3 and 1.0 mg/kg 3.0 and 10.0 mg/kg, (s.c.) 3 doses	Male Sprague–Dawley rats	300–350 g	Synaptosomes	Monoamine transporter assay microdialysis locomotion activity	Mephedrone and methylone similar to MDMA in potency and selectivity for monoamine transporters. ↑ extracellular DA and 5‐HT levels after mephedrone and methylone i.v administration in accumbens ↑ locomotor activity and stereotypy after 1 mg/kg mephedrone dose ↑ locomotor activity after 0.3 mg/kg and 1 mg/kg methylone doses.

BP = blood pressure; CPP = conditioned place preference; DA = dopamine; DAT = dopamine transporter; HR = heart rate; HPLC = high‐performance liquid chromatography; ICSS = intraccranial self‐stimulation; i.p. = intraperitoneal; i.v. = intravenous; MEPH = mephedrone; MDPV = Methylenedioxypyrovalerone; MMAI = 5‐Methoxy‐6‐methyl‐2‐aminoindane; NA = noradrenaline; NET = noradrenaline transporter; N‐methyl‐2‐AI = N‐methyl‐2‐aminoindane; PMMA = para‐Methoxy‐N‐methylamphetamine; SERT = serotonin transporter; s.c. = subcutaneous; TFMAP = CF3 substituted methcathinone analogues; USV = ultrasonic vocalisation; 3‐FMC = 3‐Fluoromethcathinone; 3‐MMC = 3‐Methylmethcathinone; 4‐MA = 4‐methylamphetamine; 4‐MEC = 4‐Methylethcathinone; 5‐HT = serotonin; 5‐IT = 5‐(2‐Aminopropyl)indole; αPBP = α‐Pyrrolidinobutiophenone; αPPP = α‐Pyrrolidinopropiophenone; αPVP = α‐Pyrrolidinovalerophenone; αPVT = α‐Pyrrolidinopentiothiophenone.

3.2. Stimulants

The most studied NPS stimulants were https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?tab=biology&ligandId=115 (3‐trifluoromethylphenylpiperazine), N‐benzylpiperazine and 4,4′‐DMAR (4,4′‐dimethylaminorex), which were all studied in 4 manuscripts (16%). The other stimulants were only studied in 3 or fewer manuscripts (Table 2). The NPS stimulants were mostly studied using neurochemical assays at dopamine, serotonin or noradrenaline systems (16 studies [64%]) with 6 studies (24%) examining toxicity or hyperthermia, 3 studies (12%) looked at the behavioural effects using models of addiction and a single study used assays related to the cardiovascular system. Thus, similar to the cathinones, the vast majority of studies examined either behavioural or neurochemical indices of addiction and very few looked at acute toxic effects. No studies examined chronic effects of stimulants and no studies reported using female rodents.

Table 2.

Preclinical studies examining stimulant novel psychoactive substance (NPS) pharmacology. The most recent manuscripts are presented first

Study	DOI	NPS	Dose/concentration	Species/sex	Age	Sample	Assay	Main results
Rickli, A. et al., 201976	10.1016/j.neuro.2019.02.011	4‐MAR, 4,4′‐DMAR, and 3,4‐DMAR				Human embryonic kidney 293 cells	Monoamine transporter assays	4,4′‐DMAR potently inhibited all monoamine transporters. 4‐MAR preferentially inhibited NET and DAT 3,4‐DMAR only weakly inhibited NET
Cai, W.T. et al., 201977	10.1016/j.neuint.2019.104487	Methiopropamine	5 mg/kg (i.p.)	Rats		Nucleus accumbens		↑number of total spine density
Maier, J. et al., 201878	10.1016/j.neuropharm.2018.06.018	4,4′‐DMAR	Dose–response curve	Human/rat		Human embryonic kidney 293 cells human striatal tissue rat pheochromocytoma cells (rPC12)	Uptake inhibition assays transporter release assays receptor and transporter binding and activation assays	↓ uptake mediated by human DAT, NET or SERT Release assays identified 4,4′‐DMAR as a substrate type releaser, capable of inducing Transporter‐mediated reverse transport via DAT, NET and SERT ↓ rat and human isoforms of VMAT2 at a potency like MDMA
Davidson, C. et al., 201879	10.3389/fpsyt.2018.00149	3,4‐CTMP and ethylphenidate	1, 10 and 100 nM/1 and 10 μM	Male Wistar rats	8 wk	Nucleus accumbens and stria terminalis	Fast cyclic voltammetry (DA and NA)	↑ evoked dopamine and noradrenaline efflux with methylphenidate (10 μM) ↑ evoked dopamine and noradrenaline efflux with 3,4‐CTMP (0.1 and 1 μM) ↑ evoked dopamine and noradrenaline efflux with ethylphenidate (1 μM)
Luethi, D. et al., 201880	10.1016/j.neuropharm.2017.08.020	N‐benzylethylphenidate, 3,4‐dichloroethylphenidate, 3,4‐dichloromethylphenidate, ethylnaphthidate, ethylphenidate, 4‐fluoromethylphenidate, isopropylphenidate, 4‐methylmethylphenidate, methylmorphenate, and propylphenidate				Human embryonic kidney 293 cells	Monoamine transporter assays	↑ inhibition NAT and DAT with all the drugs No cytotoxicity was observed after drug treatment at assay concentrations.
McLaughlin,G. et al., 201881	10.1002/dta.2396	4‐MPM/3‐MPM	Dose–response curve	Male Sprague–Dawley rats	250–300 g	Caudate (for DAT assays) or whole brain minus cerebellum and caudate (for NET and SERT assays)	Monoamine transporter assays	2‐MPM and 3‐MPM will exhibit stimulant properties like the parent compound phenmetrazine, whereas 4‐MPM may display entactogen properties more similar to MDMA.
Zwartsen, A. et al., 201882	10.1016/j.neuro.2018.03.007	BZP and TFMPP	BZP = 1–1000 μM TFMPP = 1–1000 μM	Male Wistar rats	Postnatal day 0–1	Cortical neurons culture	Multiwell microelectrodes arrays	↓ the weighted mean firing rate and weighted mean burst rate
Mayer, F.P. et al., 201883	10.1016/j.neuropharm.2017.10.006	3‐FPM, 2‐FPM and 4‐FPM	Dose–response curve	Male Sprague–Dawley rats		Human embryonic kidney 293 cells Synaptosomes	Monoamine transporter assays patch clamp	2‐, 3‐ and 4‐FPM inhibit uptake mediated by DAT and NET All FPM raised concentration‐dependent release of monoamines from rat brain synaptosomes
Sahai, M.A. et al., 201784	10.1016/j.pnpbp.2016.11.004	5‐MAPB	1,3,10 and 30 μM	Male Wistar rats	8 wk	Striatum accumbens	Radioligand binding fast cyclic voltammetry	5‐MAPB reduces the rate of dopamine reuptake 5‐MAPB binds to the DAT and displace RTI‐121 as DAT ligand
Shimshoni, J.A. et al., 201785	10.1016/j.taap.2017.01.018	MEAI	10 and 30 mg/kg	Male Sprague–Dawley rats		Striatum primary neurons human primary hepatocytes	Safety profile cytotoxic assay	Good safety profile in rats at 10 and 30 mg/kg cytotoxic effect at 500 and 1000 mg/L concentrations
McLaughlin, G. et al., 201786	10.1002/dta.2167	4F‐MPH	Dose–response curve	Male Sprague–Dawley rats	250–300 g	Synaptosomes	Monoamine transporter assays	↑ potencies determined for blockage of dopamine uptake and norepinephrine uptake In (±)‐threo isomer ↓ potent at the dopamine transporter and norepinephrine transporter in MPH
Yoon, H.S. et al., 201687	10.1016/j.bbr.2016.05.060	Methiopropamine	i.p. 0.2, 1.0, or 5.0 mg/kg	Male Sprague–Dawley rats	220–250 g		Locomotor activity	↑ sensitized locomotor activity the group that was pre‐exposed at 5 mg/kg of MPA MPA‐induced locomotor sensitization was inhibited by a preinjection of a dopamine D2 receptor antagonist
Marusich, J.A. et al., 201688	10.1016/j.neuropharm.2015.09.004	5‐IT and 6‐IT	1.0–10.0 mg/kg	Male Sprague–Dawley rats	300–400 g	Synaptosomes	Locomotor activity monoamine transporter assays	5‐IT displayed greater potency for release at DAT over SERT, while 6‐IT displayed greater potency for release at SERT over DAT 5‐IT produced locomotor stimulation
Simmler, L.D. et al., 201689	10.1124/jpet.115.229765	101 compounds: amphetamines, phenethylamines, aminoindanes, cathinones, ephedrines, tryptamines, piperazines and pipradrols	10 pM to 10 μM			Human embryonic kidney 293 cells rat and mouse TAAR1	Radioligand binding assay functional TAAR1 activity	Species differences in activity at TAAR1 among the highly active ligands, with a rank order of rat > mouse > human
Asaoka, N. et al., 201690		5F‐ADB	1 μM			Midbrain dopaminergic neurons	Electrophysiological recordings	5F‐ADB significantly increase the spontaneous firing rate in dopaminergic but nor in serotoninergic neuron
Persona, K. et al., 201691	10.1007/s12640‐016‐9604‐x	N‐Benzylpiperazine	0.1, 0.3, 1, 3, 10, 30, 100, 300, and 1000 μg/mL 10, 30, 100, and 300 μg/mL			Human glioblastoma cells (LN‐18)	Cytotoxicity assay, real‐time polymerase chain reaction gene expression analysis	↑ LDH levels ↑ mitochondrial membrane potential, ↓ ATP and ↑ reactive oxygen species production, ↑ levels of DNA damage marker (8‐OHdG) and activation of caspases: −3 and − 9.
Arbo, M.D. et al., 201692	10.1007/s00204‐016‐1665‐3	1‐benzylpiperazine, 1 (3,4methylenedioxybenzyl) piperazine, 1‐(3‐trifluoromethylphenyl) piperazine and 1‐(4‐methoxyphenyl) piperazine	625, 210 and 0.5 μM for BZP 35, 12 and 0.5 μM for TFMPP 522, 175 and 0.5 μM for MeOPP 467, 160 and 0.5 μM for MDBP	Male Wistar rats	300–400 g	Hepatocytes	Cytotoxicity assay, gene array	Key enzymes of cholesterol biosynthesis, glycoprotein transmembrane NMB and fatty acid desaturase 1 were upregulated by all 4 piperazine drugs. Betaine‐homocysteine‐S‐methyltransferase 2 were downregulated all 4 piperazine derivatives
Arbo, M.D. et al., 201693	10.1002/jat.3153	N‐ benzylpiperazine, 1‐(3‐trifluoromethylphenyl) piperazine, 1‐(4‐methoxyphenyl)piperazine and 1‐(3,4‐methylenedioxybenzyl)piperazine	500 or 1000 μM BZP 5, 50 or 100 μM TFMPP 250 or 500 μM MeOPP or MDBP			SH‐SY5Y cell	Cytotoxic assay	1‐(3‐trifluoromethylphenyl) piperazine was the most cytotoxic
Rubio, M. et al., 201594	10.1016/j.neulet.2015.01.075	4‐Methylamphetamine (4‐MA)	s.c. 2.5, 5 and 10 mg/kg	Male Sprague–Dawley rats	125–175 g		Locomotor activity, radiotelemetry	↑dose‐dependent manner, locomotor activity dose‐dependent hypothermic response to 4‐MA
Rickli, A. et al., 201595	10.1016/j.euroneuro.2014.12.012	Para‐halogenated amphetamines and pyrovalerone cathinones				Human embryonic kidney 293 cells	Monoamine transporter assays	↑ 5‐HT properties in 4‐methyl, 4‐ethyl, and 4‐bromo groups 3,4‐methylenedioxypyrovalerone, pyrovalerone, α‐pyrrolidinovalerophenone, 3,4‐methylenedioxy‐α‐pyrrolidinopropiophenone and 3,4‐methylenedioxy‐α‐pyrrolidinobutiophenone potently inhibited NET and DAT but not SERT
McLaughlin, G. et al., 201596	10.1002/dta.1732	MDMAR and DMAR	1 nM – 10 μM	Male Sprague–Dawley rats	250–300 g	Synaptosomes	Synaptosomal uptake/release	More efficacious than MDMA
Arbo, M.D. et al., 201497	10.1016/j.toxlet.2014.06.031	N‐ benzylpiperazine, 1‐(3‐trifluoromethylphenyl)piperazine, 1‐(4‐methoxyphenyl)piperazine and 1‐(3,4‐methylenedioxybenzyl)piperazine				H9c2 rat cardiac cell line	Cytotoxic assay	TFMPP seems to be the most potent cytotoxic compound
Brandt, S.D. et al., 201498	10.1002/dta.1668	4,4′‐DMAR	0.3 nM – 10 μM	Male Sprague–Dawley rats	250–300 g	Synaptosomes	Monoamine transporters assays	↑ release of DA, NA, 5‐HT. More potent than AMPH at 5‐HT
Dawson, P. et al., 201499	10.1016/j.pnpbp.2013.08.013	5‐APB	0.3–30 μM	Male Sprague–Dawley rats	8 wk	Prefrontal cortex, nucleus accumbens and caudate–putamen	Binding and voltammetry in slices. Fundus and aorta preps	Displaced both DAT and 5‐HT2 ligands and ↑ DA efflux. ↑ contraction in fundus and aorta
Davidson, C. et al., 2012100	10.1177/0269881111430733	Desoxypipradrol	1‐10 μM	Male Sprague–Dawley rats	8 wk	Nucleus accumbens	Voltammetry in slices	Bigger increase in DA vs cocaine

BZP = Benzylpiperazine; DA = dopamine; NA = noradrenaline; DAT = dopamine transporter; i.p. = intraperitoneal; MEAI = 5‐methoxy‐2‐aminoindane; NET = noradrenaline transporter; SERT = serotonin transporter; s.c. = subcutaneous; TFMPP = 3‐Trifluoromethylphenylpiperazine; 3‐FPM = 3‐Fluorophenmetrazine; 3‐MPM = 3‐methoxyphenmetrazine; 3,4‐CTMP = 3,4‐dichloromethylphenidate; 4F‐MPH = 4‐fluoromethylphenidate; 4‐MAR = 4‐Methylaminorex; 4‐MPM = 4‐Methylphenmetrazine; 4,4‐DMAR = 4,4′‐Dimethylaminorex; 5‐APB = 5‐(2‐aminopropyl)benzofuran; 5‐HT = serotonin; 5‐IT = 5‐(2‐Aminopropyl)indole; 5‐MAPB = 5‐(2‐methylaminopropyl)benzofuran.

3.3. SCRAs

There have been very few preclinical studies on SCRAs over the last 10 years with only 10 or so SCRAs examined in 8 studies. These include JWH and CP compounds while AKB48 was the most studied SCRA (3 studies [38%]). Despite the lack of studies, it is clear that the types of assay differ from both the cathinone and stimulant research. In the SCRA manuscripts, 5 studies (63%) examined cell toxicity, 5 studies (63%) examined their neuropharmacology, 1 study looked at cardiovascular effects and 1 study looked at models of addiction. Thus, in contrast to the cathinone and stimulant studies, there was only 1 study looking models of addiction and proportionally more studies examined toxicity (Table 3).

Table 3.

Preclinical studies examining synthetic cannabinoid receptor agonist novel psychoactive substance (NPS) pharmacology. The most recent manuscripts are presented first

Study	DOI	NPS	Dose/conc	Species/sex	Age	Sample	Assay	Main results
Bilel, S. et al., 2019101	10.3389/fnins.2019.01163	AKB48	0.25 mg/kg to 3 mg/kg	Male rats		Nucleus accumbens	Pharmacological and behavioural effects	↑ DA release in the nucleus accumbens shell at 0.25 mg/kg ↓ startle/PPI response at 3 mg/kg ↑ hypothermia, analgesia, and catalepsy at 3 mg/kg ↑impaired place preference and hypolocomotion at 0.5 mg/kg
Banister, S.D. et al., 2019102	10.1002/dta.2491	SCRAs 5F‐CUMYL‐PICA, 5F‐CUMYL‐PINACA and 5F‐CUMYL‐P7AICA	0.1, 0.3, 1, and 3 mg/kg (i.p.)	Male C57BL/6 J mice	21.2–29 g	Human embryonic kidney 293 cells AtT20 adenocarcinoma cells	Binding assay functional activity assay in vivo pharmacological assessment	All compounds were potent CB1 agonists 5F‐CUMYL‐P7AICA induced hypothermia
Kim, S. et al., 2019103	10.3390/molecules24163000	AKB48	0.1 to 100 μM			Human liver microsomes LLC‐PK1‐MDR1 LLC‐PK1‐mock cells HEK293 cells HEK293‐mock cells	Cytochrome P450 assays	↑inhibition of CYP3A4 and UGT1A9 activities
Kevin, R.C. et al., 2019104	10.3389/fphar.2019.00595	CUMYL‐4CN‐BINACA	0.03, 0.1, 0.3, and 1 mg/kg (i.v.)	Male C57BL/6 J mice		Mouse AtT20FlpIn neuroblastoma cells human embryonic kidney 293 cells	Radioligand binding assay receptor functional assay biotelemetry locomotor activity	Potent CB1 receptor agonist ↑ pro‐convulsant effects at 0.3 mg/kg
Domoto, M. et al., 2018105	10.1007/s00213‐018‐4933‐5	5F‐AMB	300 nM	Male and female C57BL/6 J mice	4–6 wk	Medial prefrontal cortex	Patch‐clamp	↓ excitatory and inhibitory transmission in mPFC L5 pyramidal neurons via the activation of CB1 receptors located in presynaptic terminals
De Luca, M.A. et al., 2016106	10.1016/j.neuropharm.2015.11.017	BB‐22, 5F‐PB‐22, 5F‐AKB‐48 and STS‐135	BB‐22: 0.003–0.1 mg/kg 5F‐PB‐22: 0.01 mg/kg 5F‐AKB‐48: 0.1 mg/kg STS‐135: 0.15 mg/kg	Male Sprague–Dawley rats C57BL/J6 and CB1knockout mice	200–300 g 17–20 g	Nucleus accumbens medial prefrontal cortex	Binding assay in vivo microdialysis	5F‐AKB‐48 and STS‐135 had higher Ki for CB1 binding. ↑ DA in the accumbens shell with all compounds
Yun, J. et al., 2016107	10.1039/c6tx00259e	JWH‐030	0.1, 1, 10, or 100 μM 0.001–2500 μM 30 μM i.v. 0.5 mg/kg	Male Sprague–Dawley rats male New Zealand white rabbits	7 wk 2890–3380 g	H9c2 cells rabbit Purkinje fibers	Cytotoxic assay patch clamp action potential duration (APD) electrocardiogram	JWH‐030 was more cytotoxic than JWH‐210, JWH‐250 and RCS4 JWH‐030 to block the hERG channel JWH‐030 significantly reduced the APD at 90% repolarization JWH‐030 prolonged the QT interval in rats
Tomiyama, K. 2014108	10.1016/j.taap.2013.10.028	CP‐55,940, CP‐47,497, CP‐47,497‐C8, HU‐210, JWH‐018, JWH‐210, AM‐2201, and MAM‐2201	1–30 μM for 0.5–3 h	ICR mouse	15‐d gestation	Primary neuronal cell culture	Forebrain cell cultures for caspase‐3	Apoptosis

AKB‐48 = N‐(1‐adamantyl)‐1‐pentyl‐1H‐indazole‐3‐carboxamide; AM2201 = [1‐(5‐fluoropentyl)indol‐3‐yl]‐naphthalen‐1‐ylmethanone; BB‐22 = quinolin‐8‐yl 1‐(cyclohexylmethyl)indole‐3‐carboxylate; CUMYL‐4CN‐BENICA = 1‐(4‐cyanobutyl)‐N‐(2‐phenylpropan‐2‐yl)indazole‐3‐carboxamide; CP47497 = 2‐[(1R,3S)‐3‐hydroxycyclohexyl]‐5‐(2‐methyloctan‐2‐yl)phenol; CP55940 = 2‐[(1R,2R,5R)‐5‐hydroxy‐2‐(3‐hydroxypropyl)cyclohexyl]‐5‐(2‐methyloctan‐2‐yl)phenol; DA = dopamine; HU‐210 = 1,1‐dimethylheptyl‐ 11‐hydroxytetrahydrocannabinol; JWH‐018 = naphthalen‐1‐yl‐(1‐pentylindol‐3‐yl)methanone; JWH‐030 = naphthalen‐1‐yl‐(1‐pentylpyrrol‐3‐yl)methanone; JWH‐210 = (4‐ethylnaphthalen‐1‐yl)‐(1‐pentylindol‐3‐yl)methanone; NA = noradrenaline; DAT = dopamine transporter; MAM2201 = [1‐(5‐fluoropentyl)indol‐3‐yl]‐(4‐methylnaphthalen‐1‐yl)methanone; NET = noradrenaline transporter; P7‐AICA = pyrrolo[2,3‐b]pyridine‐3‐carboxamide; SERT = serotonin transporter; STS‐135 = N‐(1‐adamantyl)‐1‐(5‐fluoropentyl)indole‐3‐carboxamide; i.p. = intraperitoneal; i.v. = intravenous; 5F‐AKB‐48 = N‐(1‐adamantyl)‐1‐(5‐fluoropentyl)indazole‐3‐carboxamide; 5F‐AMB = methyl 2‐[[1‐(5‐fluoropentyl)indazole‐3‐carbonyl]amino]‐3‐methylbutanoate; 5F‐CUMYL PICA = 1‐(5‐fluoropentyl)‐N‐(2‐phenylpropan‐2‐yl)indole‐3‐carboxamide; 5F‐PB‐22 = quinolin‐8‐yl 1‐(5‐fluoropentyl)indole‐3‐carboxylate.

We note previous studies on SCRAs from outside our search period (January 2008–December 2018), such as Randall et al.109 In this article, we focussed only on studies from the last 10 years, which include the so‐called third generation SCRAs.110, 111

3.4. Other NPS

There are relatively few studies on other NPS (20 studies) with the most studied drugs being methoxetamine and 25B‐NBOMe (both 4 studies, 20%). Most of the assays used are to examine the neuropharmacology of the drugs (12 studies [60%]) with only 1 manuscript examining effects of the drugs in models of abuse. Four studies (20%) looked at the toxicity of these drugs, 2 studies (10%) found cognitive dysfunction, 1 found bladder and renal toxicity, and 1 looked at cardiovascular effects (Table 4).

Table 4.

Preclinical studies examining other novel psychoactive substance (NPS) pharmacology. The most recent manuscripts are presented first

Study	DOI	NPS	Dose/conc	Species/sex	Age	Sample	Assay	Main results
Luethi, D. et al., 2019112	10.1016/j.ejphar.2019.05.014	2C‐BI derivatives				Human cells	Monoamine transporter assays	2C‐BI‐8 and 2C‐BI‐12 activated serotonin 5‐HT2A and 5‐HT2B receptors at sub‐μM concentrations
Wallach, J. et al., 2019113	10.1016/j.ejphar.2019.172427	Fluorolintane and its 5 aryl‐fluorine‐substituted isomers		Rats		Hippocampal slices	Monoamine transporter assays field‐recording electrophysiology Behavioural test	↑affinity for NMDA receptors ↓ long‐term potentiation ↓NMDA receptor‐induced field excitatory postsynaptic potentials ↓ PPI
Costa, G. et al., 2019114	10.1016/j.neuropharm.2018.10.031	Methoxetamine	0.1–0.5 mg/kg, i.p., × 5 d	Rats			Ultrasonic vocalizations Behavioural test	Pre‐treatment with MXE impair alterations in the elevated plus maze, marble burying and novel object ↑ neurotoxicity
Yoon, K.S. et al., 2019115	10.1007/s12012‐018‐9489‐4	Methoxetamine	0.1–500 μM	Mice		H9c2 cells	Cytotoxic assays	↓ cell viability and PAK‐1 mRNA levels at 10 μM ↓beating rate of primary cardiomyocytes at 100 μM
Halberstadt,A.L. et al., 2019116	10.1016/j.neuropharm.2018.10.037	DOB, 2C‐B and benzodifuranyl and tetrahydrobenzodifuranyl analogs		C57BL/6 J mice			Head twitch response (HTR) assay	DOB and 2C‐B induced the HTR ↑potency of DOB‐DFLY and 2C‐B‐DFLY than DOB and 2C‐B 2C‐I‐FLY, 2C‐E‐FLY and 2C‐EF‐FLY active in the HTR assay but had low potency
Herian, M. et al., 2019117	10.1007/s12640‐019‐00033‐x	25I‐NBOMe	0.3, 1, 3, and 10 mg/kg	Male Wistar‐Han rats	280–350 g	Frontal cortex	Microdialysis	Inverted U‐shaped dose–response curve on extracellular DA and 5‐HT levels U‐shaped dose–response curve on GLU levels
Zwartsen, A. et al., 201838	10.1016/j.neuro.2018.03.007	2C‐B, 25B‐NBOMe	2C‐B = 1–300 μM 25B‐NBOMe = 0.01–30 μM	Male Wistar rats	Postnatal day 0–1	Cortical neurons culture	Multi‐well microelectrodes arrays	↓ the weighted mean firing rate and weighted mean burst rate
Shintani‐Ishida, K. et al., 2018118	10.1111/1556‐4029.13583	25B‐NBOMe	i.p. 0.5 mg/kg	Male Sprague–Dawley rats	8 wk	Blood samples lung, heart, kidney or brain tissue homogenate	HPLC	↑ 25B‐NBOMe concentration in blood samples after 6 hours 25B‐NBOMe accumulated primarily in the lung
Cha, H.J. et al.,2018119	10.1016/j.neulet.2018.04.009	4‐chloro‐2,5‐dimethoxyamphetamine and AH‐7921	DOC: 0.1, 0.3, and 0.5 mg/kg AH‐7921: 0.1, 0.3, and 1 mg/kg	Male Sprague–Dawley rats C57BL/6 mice	180–220 g 15–20 g		Conditioned place preference Self‐administration	↑ preference with both drugs at 0.3 mg/kg ↑ number of responses to the active lever in the self‐administration test at 0.01 mg/kg
Luethi, D. et al., 20187	10.1016/j.ejphar.2017.12.012	Diclofensine, diphenidine, and methoxphenidine				Human embryonic kidney 293 cells	Monoamine transporter assays	Diclofensine bound to adrenergic, dopamine, serotonin and trace amine‐associated receptors. Diphenidine bound to adrenergic α1A and α2A receptors and serotonin 5‐hydroxytryptamine 1A (5‐HT1A) and 5‐HT2A receptors. Methoxphenidine bound to adrenergic α2A and serotonin 5‐HT2A and 5‐HT2C receptors
Hondebrink, L. et al., 2017120	10.1016/j.neuropharm.2017.04.035	Methoxetamine	0.1 M	Male Wistar rats	Postnatal day 1	Rat primary cortical cells Human SH‐SY5Y cells Human embryonic kidney 293 cells	Cell intracellular calcium imaging Multi‐well microelectrodes arrays Monoamine transporter assays	↑ the glutamate‐evoked increase in [Ca2+] in rat primary cortical cells with 10 μM methoxetamine ↓ the K + ‐ and acetylcholine‐evoked increase in [Ca2+]I in human SH‐SY5Y cells with 10 μM methoxetamine ↓ spontaneous neural activity between 10‐100 μM methoxetamine ↓ uptake via monoamine transporters (DAT, NET and SERT)
Kang, H. et al., 2017121	10.1016/j.neuropharm.2016.08.004	Ephenidine	Dose–response curve 1 and 10 μM	Male rats (Wistar and Sprague Dawley)	3–10 wk	Whole rat brain CA1 of rat hippocampal slices hippocampal pyramidal cells	Receptor binding assays, extracellular recording of field, excitatory postsynaptic potentials patch clamp	Ephenidine acts at the PCP site of the NMDA receptor and has lower affinity for the dopamine and noradrenaline transporters ↑ inhibition of the NMDA receptor mediated fEPSP at 10 μM 10 μM blocked NMDA receptor‐mediated EPSCs
Rickli, A. et al., 2016122	10.1016/j.euroneuro.2016.05.001	DiPT, 4‐OH‐DiPT, 4‐OH‐MET, 5‐MeO‐AMT, and 5‐MeO‐MiPT lysergic acid diethylamide, psilocin, N,N‐dimethyltryptamine and mescaline					Monoamine transporter assay	↓ binding 5‐HT2A with all the tryptamines and psilocin and DMT compared with LSD DMT, DiPT, 4‐OH‐DiPT and 4‐OH‐MET, interacted partially with the norepinephrine transporter LSD but not the tryptamines interacted with adrenergic and dopaminergic receptors
Wallach, J. et al., 2016123	10.1371/journal.Pone.0157021	DPH, 2‐MXP and 3‐ and 4‐MeO‐ isomers and 2‐cl‐diphenidine (2‐cl‐DPH)	1uM and 10uM 1.25,2.5,5 and 10 mg/kg (s.c)	Male Wistar rats male Sprague–Dawley rats	9–10 wk 250–275 g	Whole brain hippocampal slices	Binding studies, monoamine reuptake inhibition assays, in vitro field excitatory postsynaptic potential PPI	DPH and 2‐MXP, were found to be relatively selective NMDAR antagonists and inhibited NMDAR mediated field EPSPs DPH and 2‐MXP significantly inhibited PPI
Rickli, A. et al., 2015 124	10.1016/j.neuropharm.2015.08.034	2C drugs				Human cells	Monoamine transporter assay	NBOMe drugs were very potent 5‐HT2A receptor agonists 2C drugs increased the binding affinity at serotonergic 5‐HT2A, 5‐HT2C, adrenergic α1, dopaminergic D1–3, and histaminergic H1 receptors and monoamine transporters but reduced binding to 5‐HT1A receptors and TAAR1
Blough, B.E. et al., 2014125	10.1007/s00213‐014‐3557‐7	Tryptamines (21 compounds)		Rats		Synaptosomes	Monoamine transporter assay	All tryptamines were 5‐HT2A agonists. N‐ethyltryptamine was the greatest 5‐HT releaser and 5‐MeO‐MIPT was the weakest 5‐HT uptake inhibitor.
Dargan, P.I. et al., 2014126	10.3109/15563650.2014.892605	Methoxetamine	i.p 30 mg/kg x 3 mo	Mice	2–5 mo	Bladder and kidneys	CD4 and Sirius red staining	Bladder and renal toxicity
Paulke, A. et al., 2013127	10.1016/j.jep.2013.04.044	Argyreia nervosa and LSA					In silico and radioligand assays	Lower affinity than LSD but clear affinity at 5‐HT1A, 5‐HT2, and α2
Compton, D.M. et al., 2011128	10.1016/j.physbeh.2011.01.021	5‐MeO‐DIPT	6 x 5 or 20 mg/kg	Male long Evans rats	35 or 48 d		Variety of behavioural tests	Cognitive deficits

AH7921 = 3,4‐dichloro‐N‐[[1‐(dimethylamino)cyclohexyl]methyl]benzamide;hydrochloride; DA = dopamine; DAT = dopamine transporter; DiPT = N‐[2‐(1H‐indol‐3‐yl)ethyl]‐N‐propan‐2‐ylpropan‐2‐amine; DOB = 1‐(4‐bromo‐2,5‐dimethoxyphenyl)propan‐2‐amine; DPH = diphenidine (1‐(1,2‐diphenylethyl)piperidine); HPLC = high‐performance liquid chromatography; LSA = lysergic acid amide; NA = noradrenaline; NET = noradrenaline transporter; PPI = prepulse inhibition; SERT = serotonin transporter; 2C‐B = 2,5‐dimethoxy‐4‐bromophenethylamine; 25I‐NBOMe = 2‐(4‐iodo‐2,5‐dimethoxyphenyl)‐N‐[(2‐methoxyphenyl)methyl]ethanamine; 2‐MXP = 2‐methoxydiphenidine (1‐[1‐(2‐methoxyphenyl)‐2‐phenylethyl]piperidine); 4HO‐DiPT = 4‐Hydroxy‐N,N‐diisopropyltryptamine; 4HO‐MET = 4‐hydroxy‐N‐methyl‐N‐ethyltryptamine; 5‐MeO‐AMT = 1‐(5‐methoxy‐1H‐indol‐3‐yl)propan‐2‐amine; 5‐MeO‐DiPT = N‐[2‐(5‐methoxy‐1H‐indol‐3‐yl)ethyl]‐N‐propan‐2‐ylpropan‐2‐amine; 5‐MeO‐MiPT = N‐[2‐(5‐methoxy‐1H‐indol‐3‐yl)ethyl]‐N‐methylpropan‐2‐amine.

3.5. Overview of preclinical data on NPS (more formal and supported)

Taken together we can see that most recent studies on NPS examined the reinforcing effects or addictive liability of NPS, whether through behavioural models (32%) or neurotransmitter changes such as dopamine levels (46%). Taking NPS for their rewarding effects is a motivation for use; other reasons for using NPS including self‐medicating to treat withdrawal symptoms from other drugs or the availability when traditional drugs are scarce or supplementing the illicit drug use by trying to induce synergistic or additive effects. From a legislative point of view, it might be useful to know if a drug was more or less addictive and this criterion is used by the EU for banning NPS.129 Further, more resources, whether it be policing, research, healthcare or education could be targeted at those NPS predicted to be more addictive. The next most common type of study examined cell toxicity (24%). This is useful as highly toxic substances should clearly be banned and this criterion is also used by the EU. Interestingly, the effects of these drugs, which might result in acute hospitalisation, for example cardiotoxicity, seizure activity, hyperthermia or death, have hardly been studied at all in NPS (all <5%).

Only 4% of studies included female rodent samples and only 9% included chronic dosing studies. The age range of most studies was appropriate with most (72%) using rodents, which were adolescent or in early adulthood (100–299 g) and only 22% of studies examining older rodents (>300 g).

4. PREDICTIVE VALIDITY OF NPS TESTS IN ANIMAL TISSUE

As detailed above, when NPS started flooding the market over the last decade or so, preclinical tests were exploited to compare these substances to their traditional counterpart, describing them as https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=2286 ‐like or amphetamine‐like substances.60, 130 Unlike known illicit substances such as https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=4803 and MDMA (3,4‐methylenedioxymethamphetamine), the pharmacology related to emerging NPS is scarce and may not suitably be extrapolated from existing knowledge. For example, when the risk assessment of 2C‐I (2,5‐dimethoxy‐4‐iodophenethylamine) was carried out in 2003, a speculative comparison was made with the phenethylamine analogue 2C‐B (2,5‐dimethoxy‐4‐bromophenethylamine) and the amphetamine analogue DOB (https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=163), as they both have similar chemical structures to 2C‐I but with a bromine instead of iodine. It was deemed inappropriate to compare or extrapolate data related to the structurally similar MDMA, PMA (paramethoxyamphetamine) and 4‐MTA (4‐methylthioamphetamine) due to the absence of the 2,5‐dimethoxy functional group.131

The validity of animal models in understanding the harmful effects of illicit substances in humans is well established and therefore, can provide a predictive validity for NPS‐related harms. For instance, ketamine has been associated with bladder toxicity.132 Similar to ketamine, animal models showed significant bladder and renal toxicity in rodents following the administration of 30 mg/kg methoxetamine intraperitoneally.126 Symptoms included inflammatory cell infiltration, tubular cell necrosis, glomerular damage and increased micturition frequency bladder dysfunction.126 ^, 133

Benzofurans such as 5‐APB (5‐(2‐aminopropyl)benzofuran), 5‐APDB (5‐(2‐aminopropyl)‐2,3‐dihydrobenzofuran), 6‐APB (6‐(2‐aminopropyl)benzofuran), 6‐APDB (5‐(2‐aminopropyl)‐2,3‐dihydrobenzofuran), 4‐APB (4‐(2‐aminopropyl)benzofuran), 7‐APB (7‐(2‐aminopropyl)benzofuran), 5‐EAPB (5‐(2‐ethylaminopropyl)benzofuran) and 5‐MAPDB (1‐(2,3‐dihydrobenzofuran‐5‐yl)‐N‐methylpropan‐2‐amine) are all structurally similar to MDMA.134 Similar to MDMA, they were found to activate the 5‐HT_2B receptor (Dawson et al., 2014), which induces heart valve fibrosis124, 135, 136 and inhibit dopamine transporters (DAT).84, 99

MDPV (3,4‐methylanedioxypyrovalerone) is a cathinone derivative with a nitrogen atom in the pyrrolidine ring and a 3,4‐methylenedioxy group on the phenyl ring similar to MDMA.137 in vitro and in vivo rodent models showed that MDPV blocks the dopamine and norepinephrine transporters in a similar way to the pyrovalerone analogues. It is more potent at both the dopamine and norepinephrine transporters and less potent blocking serotonin in a similar way to cocaine.137, 138 The potent blockade of dopamine and epinephrine stipulates that MDPV has the potential of inducing a high risk of abuse/addiction, and life‐threatening cardiovascular stimulation including tachycardia and hypertension respectively, more than cocaine.137, 138 This was consistent with in vitro data assessing blood–brain barrier permeability, in vivo microdialysis and in vivo locomotor activity testing in rats.130, 137, 138

One problem with preclinical studies is whether or not the toxic dose or concentration from the preclinical study reflects clinical doses. There is a wealth of literature on MDMA toxicity in the 5‐HT system, but this has been criticised because the doses needed to show neurotoxicity in animals may be far above the clinically relevant doses.139

Taken together, we can see that preclinical studies can be used to predict health harms. Some researchers have gone a step further, using in silico studies to predict addictive liability of NPS. The benzofurans 5‐APB and 5‐MAPB were predicted to bind to DAT in a similar way to MDMA,84 while ketamine‐like NPS diphenidine and methoxphenidine were predicted to bind differently at DAT.140

5. EVIDENCE USED IN UK LEGISLATION AND IN EUROPE

5.1. Evidence used in the UK

In the UK, the government takes advice on issues around drugs of abuse from the ACMD who risk assess emerging substances and issue subsequent recommendations.141 The ACMD also advises the government on the control of drugs and drugs requiring a temporary class drug order. In their first NPS report, on BZP (benzylpiperazine), they recommended that the drug and some of its 1‐phenyl and 1‐benzyl derivatives be brought under the Misuse of Drugs Act (1971): “The ACMD considers that the harms and misuse of BZP and substituted piperazines (identified in Annex 4) are commensurate with Class C, under schedule 2, part III, of the Misuse of Drugs Act (1971); and should be scheduled under Schedule I of the Misuse of Drugs Regulations (2001) (having no recognised medicinal use).” This decision was based on a report by the EMCDDA.142 In this report there is evidence from preclinical studies in the 1970s and 80s that BZP had addictive liability and had similar properties to amphetamine or MDMA. These studies were undertaken as the drug was being examined for antidepressant effects. More recent data from the 2000s are focused on human use and in particular BZP use in New Zealand where it was a popular party drug. These data suggest that the drug should not be used by people susceptible to seizures or in those with cardiac conditions. It is difficult to attribute blame to a particular drug in many of the human studies as in most cases the drug takers, although being shown to have BZP in their systems, may have consumed numerous other substances. Thus, although the UK government banned BZP, as it was obliged to do by the European Commission, the direct evidence for toxicity was limited because much of the data came from polydrug users. Nevertheless, its known pharmacological profile and its suspected effects on humans clearly merited some sort of control.

In another early publication, this time advice to the then Home Secretary Theresa May, on D2PM (Ivory wave) which “… typically cause prolonged agitation (lasting up to 5 days after drug use which is sometimes severe, requiring physical restraint), paranoia, hallucinations and myoclonus (muscle spasms/twitches)”, the ACMD recommended that there was an immediate ban on the import of 2‐DPMP. This advice was followed a year later by a report on desoxypipradrol,3 the ACMD could again rely on data from previous studies as desoxypipradrol had been tested for narcolepsy. After considering information from hospital emergency rooms, published research, drug company data and coroners reports the advice was to bring desoxypipradrol and its associated drugs D2PM and 2‐diphenylmethylpyrrolidine under the Misuse of Drugs Act (1971). The clinical evidence used to ban desoxypipradrol included 3 studies: a case report of a man who had taken D2PM with agitation and chest pains143; 5 case reports have shown the presence of D2PM rather than desoxypipradrol and the victims exhibited signs of agitation, anxiety and insomnia but not any sympathomimetic toxicity144; 2 case reports involving polydrug abusers, again presented with agitation, anxiety and insomnia but without increased heart rate, hypertension or hyperthermia.145 None of these studies actually confirmed desoxypipradrol use. Recent preclinical data have shown that desoxypipradrol was more potent than cocaine at inhibiting dopamine reuptake.100 Perhaps the best evidence to ban desoxypipradrol came from a report by Ciba‐Geigy in the 1950s showing that desoxypipradrol tended to have a much lower 50% lethal dose in animals than amphetamine or methamphetamine.

The ACMD went on to provide a number of reports and advice on various NPS including cathinones (2010), methoxetamine (2012), benzofurans (2013), N‐BOMe, AH‐7921 (3,4‐dochloro‐N‐[[1‐(dimethylamino)cyclohexyl]‐benzamide), tryptamines, MT‐45 (1‐cyclohexyl‐4‐(1,2‐diphenylethyl)‐piperazine) and 4,4'DMAR (all 2014), MPA (methiopropamine; 2016), third generation SCRAs (2016), https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=7236‐like NPS (2017) and 2–4‐dinitrophenol (2019). Not all NPS had a previous history as a potential medicine, where there were previous preclinical and clinical studies on the compound. For example, see the EMCDDA risk assessments where the pharmacodynamic profiles of emerging substances such as SCRAs on other pharmacological targets other than CB₁ is very limited. This is particularly true for AB‐CHMINACA (N‐(1‐amino‐3‐methyl‐1‐oxobutan‐2‐yl)‐1‐(cyclohexylmethyl)‐1H‐indazole‐3‐carboxamide), ADB‐CHMINACA (N‐(1‐amino‐3,3‐dimethyl‐1‐oxobutan‐2‐yl)‐1‐(cyclohexylmethyl)‐1H‐indazole‐3‐carboxamide), 5F‐MDMB‐PINACA (5F‐ADB or methyl‐2‐{[1‐(5‐fluoropentyl)‐1H‐indazole3‐carbonyl]amino}‐3,3‐dimethylbutanoate), CUMYL‐4CN‐BINACA (1‐(4‐cyanobutyl)‐N‐(2‐phenylpropan‐2‐yl)indazole‐3‐carboxamide) and many more.

In May 2016, the UK government adopted a different approach aimed at limiting the supply of NPS and capturing substances that escaped the Misuse of Drugs Act 19714 rather than on evidence‐based risk assessments. However, as explained above, implementation of the Act is highly reliant on in vitro databases of a representative set of molecules.6 This is quite limited given the time taken to develop these libraries and the limitations of these libraries when facing emerging previously unseen substances.

Take together we can see that preclinical evidence has previously played an important role in deciding to ban NPS in the UK and continues to support the implementation of the new legislation. The clinical evidence, mostly case reports, is somewhat unreliable due to polydrug use and comorbidity, amongst other issues. Much of the preclinical evidence only suggested potential harms, rather than showing actual toxicity, and is dogged by the obvious need to extrapolate from animals to humans.

5.2. Overview of evidence used in Europe

In Europe, the process has been quite similar to that described above for the UK. The EMCDDA have published numerous reports and advice on NPS. These started with BZP (2007, 2009), Spice (2009), 4‐MA (2012), MDPV, methoxetamine, AH7921, 25I‐NBOMe, 4,4′‐DMAR and MT‐45 (2014) and more recently several reports on SCRAs and synthetic opioids. In their report129 on “New Psychoactive Substances in Europe; Legislation and Prosecution; Current challenges and Solutions” they describe 4 broad attempts of member countries controlling NPS. First, by trying to control NPS under laws around medicinal products, but this was thrown out by the Centre for Justice for the European Union on the grounds that NPS were not medicines. Second, some countries have tried to use existing laws around existing consumer safety laws. Third, existing drug laws have been modified by using group definitions of some NPS. Fourth, new laws have been developed, such as the UK Psychoactive Substances Act 2016 described above. In these new laws, the criteria used to define a psychoactive substance are often different and most use an element around harm or threat to health (including dependence) and a criterion that the drug has a psychoactive effect. Clearly, without preclinical data, and preferably human data, it is difficult to say with any certainty that an NPS has a psychoactive effect or could cause harm. Some countries also define psychoactivity quite clearly in their legislation with Ireland requiring significant mental disturbance or change and Hungary and Portugal requiring the NPS to have “effects similar to established drugs of abuse and a likelihood to cause dependence”.129

The EMCDDA have published 22 risk assessments for NPS. The most recent assessment will be used as an example. This was a risk assessment of the synthetic opioid cyclopropylfentanyl. The EMCDDA and Europol examined the available information based upon the following criteria: (i) the amount of the material seized; (ii) evidence of organised crime involvement; (iii) evidence of international trafficking; (iv) analogy with better‐studied compounds; (v) evidence of the potential for further (rapid) spread; and (vi) evidence of cases of serious intoxication or fatalities. They reported a pharmacological description of a few studies mostly limited to examining the effects of this NPS at μ‐opioid receptors in vitro, where it was shown to be more potent than https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=1627 or https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=1626 146 and a single animal study, in mice, suggesting analgesic properties.147 The health risks were suggested to come from accidental overdose and no acute or chronic toxicity studies had been carried out, nor had its dependence liability been examined. It was assumed that this NPS would have toxicity similar to morphine or fentanyl. The report described deaths associated with cyclopropylfentanyl, which included 78 in Sweden, 3 in the UK, 1 in Norway and over 100 in the USA,146 although in nearly all cases users had recently used multiple drugs. The report was submitted to the European Commission and Council of the European Union, who decided that cyclopropylfentanyl should be subject to control measures across member states. Thus, the EU banned this NPS based on very little direct pharmacological or toxicology data and include information on criminality as a criterion to ban the substance.

6. CONCLUSIONS

Clearly it would be beneficial to know as much as possible about each NPS, however scientific curiosity needs to be tempered by the reality of the current drug scene where the vast majority of problems are caused by relatively few established drugs of abuse. It is possible that some NPS may join these established drugs of abuse as major players in the morbidity and mortality associated with recreational drug use and it is these most commonly used NPS that we should focus upon. However, horizon scanning for the next MCAT/mephedrone or Spice/SCRA is also important. We have summarised the preclinical data on NPS above and given examples of these data being used in UK and EU legislation. It is reasonable to ask if the changes in legislation have been evidence‐based? Nutt and colleagues have published in this area for established drugs of abuse and concluded that UK and Australian scheduling of drugs of abuse is not particularly evidence based. They highlight the easy access to nicotine and alcohol, despite these drugs being ranked as highly dangerous. By contrast, ecstasy, https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=17 and magic mushrooms, despite being ranked very low on the danger scale, are schedule 1, class A drugs in the UK.148 It is difficult to say that the UK blanket ban4 on psychoactive substances is evidenced based. The vast majority of NPS have not been examined at all while those that have been examined, as described above, tend to be looked at for their addictive effects rather than toxic effects. In addition, studies examining the long‐term effects of NPS are very rare and these types of studies would be best placed from which to extrapolate health harms. Nevertheless, the ban appears to be having some of the desired effects as recent evidence suggests fewer problems associated with NPS.5 The ACMD and EMCDDA reports are generally evidence based but limited to a relatively small number of NPS or families of NPS and a relatively small number of supporting studies. Having said that, given the very large number of NPS (up to 4000 by some estimations,2 there is little chance of all of these drugs being looked at in detail. The best that we can realistically hope for is putting many of these drugs through some high throughput screens for receptor/transporter binding120, 149 and high throughput toxicity assays.150 An alternative approach is in silico testingi.e. modelling the effects of the drugs at receptors or transporters of interest. We have done this for some ketamine‐like NPS including diphenidine, methoxphenidine, MDMA, and some similar NPS including 5‐APB and 5‐MAPB.84, 140 We also suggest that future studies focus more on measures of toxicity (e.g. neurotoxicity, vasoconstriction, hepatotoxicity, seizures and hyperthermia) as this has clearly been overlooked with most studies focussing on rodent behavioural or neurochemical (dopamine) measures of dependence. Finally, there are very few preclinical studies using tissues from female rodents. This is important as drug effects do show sexual differentiation.151 Thus, in order to get an accurate overview of NPS pharmacology, more studies using female tissue are needed.

COMPETING INTERESTS

There are no competing interests to declare.

Santos‐Toscano R, Guirguis A, Davidson C. How preclinical studies have influenced novel psychoactive substance legislation in the UK and Europe. Br J Clin Pharmacol. 2020;86:452–481. 10.1111/bcp.14224