Four cases of death involving the novel synthetic cannabinoid 5F-Cumyl-PEGACLONE

On the market for new psychoactive substances (NPS), synthetic cannabinoids (SCs) represent a class of designer drugs, which have been synthesized for recreational purposes to mimic the effects of delta-9-tetrahydrocannabinol (THC). Thus, they are usually sold as ‘synthetic marijuana’ or ‘fake weed’ and claimed to be safe and legal alternatives to common illegal cannabis-based products, despite being potent full agonists at the cannabinoid CB₁-receptor.

SCs are not only one of the most widespread classes of NPS, but also among the most structurally diverse. Despite the evolution of the different national and/or international legislations, novel compounds are synthesized and sold to circumvent the law and although the number of new compounds has dropped in the last years, their availability is still high [1].

Several different classes of SCs have been synthesized since the first appearance in 2008 [2, 3], including substances characterized by indole, indazole, or benzimidazole core structures [4]. Gamma-carbolinone derived SCs are characterized by a tricyclic core structure and first entered the market in Germany in November 2016, after the “Act to control the distribution of new psychoactive substances” (NpSG), which did not cover this type of structures until recently, came into force [5].

Particularly, 5-pentyl-2-(2-phenylpropan-2-yl)-2,5-dihydro-1H-pyrido[4,3-b]indol-1-one (semisystematic name: Cumyl-PEGACLONE or SGT-151) was the first SC with a γ-carbolinone core structure detected in forensic casework and, since then, it has dominated the German SC-market [6-8]. It has indeed been detected in about 25% of herbal blends purchased in Germany during a monitoring program and has showed high prevalence in biological specimens [8].

Cases of deaths due or related to SCs are commonly reported in the literature, although the determination of the significance of toxicological results, as highlighted by Elliott et al. [9], can be difficult and often lacks. The Toxicological Significance Score (TSS) is a system developed to perform and strengthen the health risk assessment of NPS and to allow the clarification of the role of such substance(s) in deaths. On the basis of a comprehensive analysis, a score from 1 (i.e. Low, alternative cause of death) to 3 (i.e. High, NPS likely to have contributed to toxicity/death, despite the detection of other drugs) is assigned [9]. Six cases of death related to the intake of Cumyl-PEGACLONE were reported, in none of which, according to the TSS, the substance seemed to have played a causative role [8]. Due to the relatively low occurrence of severe intoxications and of deaths caused by Cumyl-PEGACLONE, the compound has been suggested as a “comparatively safe drug” and a relative “low toxicity” of SCs bearing a γ-carbolinone core structure was considered possible [8].

In 2018, following the introduction of Cumyl-PEGACLONE into the annex of the German Narcotics Law (Betäubungsmittelgesetz, BtMG), the fluorinated analog of Cumyl-PEGACLONE, 5F-Cumyl-PEGACLONE (5-(5-fluoropentyl)-2-(1-methyl-1-phenylethyl)-1H-pyrido[4,3-b]indol-1-one) was identified in our laboratory as the constituent of a “research chemical” and its metabolism has been studied by means of in vitro and in vivo studies [10]. Since then, it almost replaced Cumyl-PEGACLONE on the market, with a proportion of about 25% of the herbal blends analyzed in our laboratory, and was recently added to the list of controlled substances in Germany. In Fig. 1, Cumyl-PEGACLONE and its 5-fluoro analog are shown.

The binding affinity of 5F-Cumyl-PEGACLONE to the CB₁ receptors has not been yet studied, though a higher potency than Cumyl-PEGACLONE (1.37 ± 0.24 nM) [5] would be in line with fluorinated compounds usually being more potent than their non-fluorinated analogs [11].

Between the last months of 2018 and the first months of 2019, four cases of death, in which 5F-Cumyl-PEGACLONE was involved, were discovered in Germany. To the best of our knowledge, these are the first cases connected to this novel SC.

The standard addition method for the quantification of 5F-Cumyl-PEGACLONE was applied to postmortem samples of cases 1–4 showing a good linearity for all specimens (mean R² = 0.99). Calculated x-intercepts, range of the calibration curve, equations, goodness of fit and number of data points are given in Table 1.

Full results of the toxicological analyses for femoral blood, heart blood, urine and hair, including general screening and quantification of other drugs for cases 1–4 are shown in Table 2.

As for the validation of the LC–MS/MS method in serum, no interferences occurred in the selectivity tests. Good linearity was shown in serum for the concentration range of 0.01–2.0 ng/mL and a linear calibration model was accepted, with no need for a weighting factor (correlation coefficients (R²) = 0.997).

Data on LOD, LOQ, accuracy, precision, and matrix effects for the LC–MS/MS method (validated for serum only) used for preliminary analysis are provided in Table 3.

5F-Cumyl-PEGACLONE has only recently emerged on the German drug market, though it seems to have gained popularity and almost completely replaced Cumyl-PEGACLONE, the first γ-carbolinone-type cannabinoid that was introduced in the end of 2016. Indeed, it was detected as the only active ingredient in 12 out of 15 herbal products purchased on the Internet and analyzed in 2019 [7]. Despite its widespread consumption, pharmacokinetic and pharmacodynamic data are still limited. The compound can presumably cause typical adverse effects, such as nausea, vomiting, somnolence, dizziness, anxiety, panic attacks, psychosis, tachycardia, seizures, and comatose states [18, 19], similarly to other potent CB₁ receptor agonists. Seizures and collapse were reported in a case of SGT-151 administration [20], while in a recent study by Kevin RC et al., 2019, it was found that SCs, including those bearing a cumyl-substituent, such as 1-pentyl-N-(2-phenylpropan-2-yl)-1H-indole-3-carboxamide (CUMYL-PICA) and 1-(5-fluoropentyl)-N-(2-phenylpropan-2-yl)-1H-indole-3-carboxamide (5F-CUMYL-PICA), undergo thermal degradation with release of potentially toxic substances [21].

Several death cases involving SCs, e.g. methyl 2-{[1-(5-fluoropentyl)-1H-indazole-3-carbonyl]amino}-3,3-dimethylbutanoate (5F-ADB), N-[(1S)-1-(aminocarbonyl)-2-methylpropyl]-1-(cyclohexylmethyl)-1H-indazole-3-carboxamide (AB-CHMINACA) and [1-(5-fluoropentyl)-1H-indol-3-yl](naphthalene-1-yl)methanone (AM-2201), have been reported in the literature, often assigning a causative or contributory role to the compound(s) [11, 22-25]. The wide range of postmortem concentrations, overlapping those of non-fatal intoxications, strongly limits the possibility of identifying so-called “toxic ranges” [25, 26]. Indeed, the interpretative weight assigned to SCs in cases of death can largely vary among different evaluators [27]. The lack of data regarding postmortem redistribution and toxic effects further represents a challenge when assessing the TSS [9] of SCs in death cases and highlights that concentrations should be interpreted with caution, in a multi-disciplinary evaluation including circumstantial data and morphological assessment [9, 25].

In the cases here presented, non-specific findings of intoxication, e.g. vomiting and/or gastric aspiration, pulmonary and brain edema, injection marks, hemolysis and blood stasis in the internal organs, as well as an accumulation of urine in the bladder, were seen. The unavailability of morphological findings that could serve as cause of death, in combination with a past history of illicit drug consumption, highlighted the need for toxicological analyses. In all deaths, the detection of 5F-Cumyl-PEGACLONE in blood and the presence of its main metabolites in urine indicated a recent intake of the substance.

In case 1, the appearance of the inner myocardial layers and the aspiration of gastric content were considered as signs of an agonal phase. Skin erythema and acute gastric erosions (i.e. Wischnewski-sign) are well-known signs of hypothermia, the latter being present in 40–90% of hypothermia-related deaths [28, 29]. Notwithstanding these typical findings, a death due to hypothermia, which remains a diagnosis of exclusion, cannot be confirmed, since the victim, a well-nourished individual, was found partially dressed and indoor at his home. Moreover, neither any open window or door nor any abnormalities in the room temperature, which was not measured, were reported from the DSI. On the other hand, according to experimental evidence in rats, SCs can induce a decrease in body temperature [30] and this could be additionally hypothesized in the present case, although available clinical data of survived SC intoxications does not support this.

The main active metabolite of risperidone, 9-hydroxy-risperidone (9-OH-risperidone), also available as a drug in its own right (paliperidone), trimipramine and diphenhydramine, were each within their typical therapeutic ranges (2–20 ng/mL, 10–300 ng/mL, and 50–100 ng/mL) [31, 32]. Even if a prolonged agony time could lead to a limited reduction of the initial drug concentrations and despite the possibility of an additional decrease of the blood concentrations due to postmortem redistribution (e.g. [33].), a primary role of these drugs in the death is very unlikely. Several indazole and γ-carbolinone based SCs and their metabolites were detected in scalp hair. Due to longitudinal diffusion, delayed incorporation, thermal instability of the compounds, and several other factors, hair analyses neither allow to determine the time of SC consumption by segmental hair analysis nor to certainly discriminate active use from passive exposure, i.e. contamination of hair from the outside during smoking [34-36]. The relatively high hair concentrations, as well as the multiplicity of the detected SCs, confirmed a pattern of consumption characterized by a long-term history of SCs use and, thus, suggested a certain tolerance to their effects. This is of utter importance, especially when the blood concentrations, as the ones retrieved in the present case (0.07 ng/mL in central and 0.45 ng/mL in peripheral blood), are not as high as in other SC involving intoxication cases found in the literature [25, 37, 38]. Lower levels were detected in heart blood, despite gastric regurgitation, which can lead to an increase in cardiac concentrations in the case of oral intake. The central compartment is more prone to passive postmortem redistribution from highly concentrated organs (e.g. lungs, liver) than the peripheral one, which only relies on diffusion from adipose tissues and muscles [39, 40]. The discrepancy observed can be due to several factors, one of them being a relatively short time interval between last consumption of the drug and death. In another case presented by Vogt et al. [41] with circumstantial evidence of methyl 2-[[1-(cyclohexylmethyl)-1H-indole-3-carbonyl] amino]-3,3-dimethylbutanoate (MDMB-CHMICA) consumption shortly prior to death, similar findings were reported. This points toward the possibility of higher concentrations in peripheral blood being the consequence of incomplete primary distribution of the drug at the time of death with initially very high concentrations, which might have declined at different speed in the heart vessels and the peripheral vessels. In cases 2–4, similar concentrations in peripheral and central blood or even higher concentrations in heart blood were detected. This further suggests that the discrepancy between central and peripheral blood concentrations in case 1 is not due to a passive redistribution from surrounding tissues to peripheral blood.

Given the absence of other plausible or potential causes of death, the consumption of 5F-Cumyl-PEGACLONE as the cause of death seemed likely (a TSS of 3 was assigned), even though the exact mechanism remained unclear.

Despite hypothermia can lead to hemorrhages in pancreas, soft tissues and in synovial membranes, kidney bleedings have never been described in connection to this condition, not even when cases of death due to low temperatures were studied by means of postmortem computed tomography [28, 42-44]. In 2013, the Centers for Disease Control and Prevention (CDC) reported several cases of acute kidney injury (AKI) associated with SCs and particularly with [1-(5-fluoropentyl)-1H-indol-3-yl](2,2,3,3-tetramethylcyclopropyl)methanone (XLR-11), a fluorinated compound [45]. None of these cases led to the death of the victim despite requiring corticosteroids therapy and/or hemodialysis. Thus, an extensive histopathological analysis did not take place and only renal biopsies were performed, showing signs of acute tubular injury [45]. These findings were confirmed by histology in our case. Kidney bleedings have never been previously demonstrated in such contexts, even though the ultrasound appearance of the cases presented by the CDC, which displayed an increase in cortical echogenicity, is consistent with subcapsular hemorrhage. Moreover, in all the cases described by the CDC and by other authors, the victims suffered from nausea, vomiting, and abdominal pain, which were the chief symptoms at presentation [45-48]. AKI was also described in combination with fatal hepatic failure after the intake of another fluoro-SC, quinolin-8-yl 1-(5-fluoropentyl)-1H-indole-3-carboxylate (5F-PB-22) [38]. Indeed, it seems unlikely to relate the subacute death of victim 1 to AKI, though a harmful effect of the fluorinated compound on kidneys appears probable. Recently, a painless macro-hematuria with altered coagulation was described in association with SC use, though no anticoagulant poisoning panel was performed [47]. Additionally, there were reports about products containing SCs allegedly laced with potent anticoagulant coumarin derivatives like brodifacoum and difenacoum [48, 49]. In our case, to verify if the kidney bleedings could have been caused by such contaminants, the analysis of coumarin derivatives including super-warfarin-type drugs was performed, but was found negative.

Finally, as a mechanism of death, we suggest a cardiac toxicity of 5F-Cumyl-PEGACLONE with induction of tachycardia/arrhythmia.

In case 2, the toxicological analyses retrieved morphine (297 ng/mL), 6-acetylmorphine (6-AM, 20 ng/mL), codeine (21 ng/mL), oxazepam (450 ng/mL), and 5F-Cumyl-PEGACLONE (0.21 and 0.23 ng/mL in central and femoral blood). Heroin and morphine are sedative-hypnotic drugs, which can lead to respiratory depression. Since heroin is quickly deacetylated to 6-AM after ingestion and further metabolized to morphine, the presence of 6-AM in the blood of the deceased and the morphine/codeine ratio > 1 confirmed the intake of illicit heroin and suggested that the death occurred at a time close to the drug consumption [50-52]. The relatively low concentrations of morphine in bile (1590 ng/mL), compared to those retrieved in the literature [53], support a non-regular abuse of heroin in victim 2 [54, 55]. This finding was not consistent with the circumstantial data and with the multiple signs of past drug abuse retrieved at the autopsy (i.e. scars and crusts in injection sites and poor condition of the dentition in a woman of only 48 years). Even in the presence of past illicit drug use, it cannot be excluded that the victim had experienced a drug-free period and, thus, presented a low tolerance to the harmful effects of heroin, the main active metabolite of which was found in a concentration well above the typical toxic threshold (100 ng/mL for morphine [32, 56]).

It is to be noted that no synthetic opioids (e.g. fentanyl derivatives) were detected, while the consumption of several SCs was noted.

Despite the measured concentrations of 5F-Cumyl-PEGACLONE were relatively low compared to opiates, similar levels were found in the other cases and even lower levels were retrieved in femoral blood of cases 3 and 4. Thus, such concentrations cannot be neglected.

The detection of metabolites of methyl (2S)-2-{[1-(4-fluorobutyl)-1H-indazole-3-carbonyl]amino}-3,3-dimethylbutanoate (4F-MDMB-BINACA) and of a common metabolite of N-(1-amino-3-methyl-1-oxobutan-2-yl)-1-(4-fluorobenzyl)-1H-indazole-3-carboxamide (AB-FUBINACA) and methyl (2S)-2-[[1-[(4-fluorophenyl)methyl]indazole-3-carbonyl]amino]-3-methylbutanoate (FUB-AMB) in urine but not in blood proves that the consumption of these drugs occurred in a larger time distance from death and was not relevant. Finally, oxazepam is a short-to-intermediate-acting benzodiazepine also known as an oxidative metabolite of other benzodiazepines and displays anxiolytic and sedative effects. Benzodiazepines are relatively safe drugs and, in the present case, oxazepam and alprazolam were both within the typical therapeutic ranges, so that a significant role in the death seems unlikely.

In the light of the above-mentioned, the death was attributed to an acute heroin intoxication. However, 5F-Cumyl-PEGACLONE may be considered as an additional factor in the course of the intoxication. As heroin definitely was the decisive intoxicant, a toxicological significance score of 1 was assigned to 5F-Cumyl-PEGACLONE.

In case 3, non-toxic concentrations of temazepam (230 ng/mL), oxazepam (12 ng/mL, probably found as a metabolite of temazepam), and alprazolam (16 ng/mL) were detected [32]. Moreover, pregabalin (6000 ng/mL), an antiepileptic drug also used to treat anxiety disorders, was detected in both heart and femoral blood. Patients treated with 600 mg/day of pregabalin usually show plasma concentrations between 2000 and 8000 ng/mL [56], with upper levels reaching 14,000 ng/mL [57]. Postmortem blood concentrations probably being causative to death were reported to be higher than 25,000 ng/mL [58]. As the victim was allegedly under constant temazepam and pregabalin medication, toxicological analyses confirmed the circumstantial data.

Concerning 5F-Cumyl-PEGACLONE, concentrations in central and peripheral blood were 0.22 ng/mL and 0.12 ng/mL, respectively. In this case, the attribution of a significance score (TSS) of 2 (“Medium”) to 5F-Cumyl-PEGACLONE is mainly due to the absence of competing causes of death at the postmortem examination and to the possible contributory role of pregabalin and temazepam (despite determined within therapeutic concentrations), since these compounds might have lowered the threshold for a central nervous system depressant effect or have interacted with the SC.

A SC intake was strongly suggested by the DSI, which allowed the identification of a herbal blend later confirmed to contain 5F-Cumyl-PEGACLONE. The mechanism of death remained unclear.

In case 4, the cocaine metabolites benzoylecgonine and ecgonine methyl ester as well as 11-nor-9-carboxy-Δ9-tetrahydrocannnabinol (THC-COOH), an oxidative metabolite of THC, were detected in blood, in the absence of the parent compounds. Cocaine is a centrally stimulating drug potentially leading to cardiac arrhythmias, an increase in body temperature, cerebral seizures, and respiratory paralysis, that is rapidly degraded in blood (half-life of 0.5–1.5 h [59]) resulting in a reduction of postmortem levels [60]. Thus, toxicity can hardly be estimated from blood concentrations [61]. In the present case, the absence of parent compound and the relatively low concentrations of the ester hydrolysis products indicate that the last cocaine intake was either very low-dosed or occurred with a larger time distance to death. Moreover, case 4 did not fit to the typical picture of cocaine-related fatalities [62], since no signs of (sub) acute cardiac events, aortic dissection or ruptured aneurysm were seen. The trace finding of THC-COOH, which is not pharmacologically active, was regarded as irrelevant. The aspiration of gastric content in a healthy, young man contradicts the assumption of a sudden death and/or a drug-induced arrhythmia, and is most probably secondary to a loss of consciousness. The big amount of food in the stomach and in the chest could be due to a “food-craving” effect, which is commonly experienced in association with cannabis [63, 64], though the appetite stimulation was reported less intense with SCs [65]. Somnolence and lethargy are commonly reported in association with SCs [66] and there are no other findings which could explain the state of low consciousness experienced by victim 4.

Hair analysis revealed both 5F-Cumyl-PEGACLONE and Cumyl-PEGACLONE. It is known that the presence of SCs in hair samples can arise not only from active intake, but also from an external contamination [13]. However, considering the relatively high concentrations of 5F-Cumyl-PEGACLONE and the co-occurrence of Cumyl-PEGACLONE, a history of SC consumption in this case is likely.

A prominent role of 5F-Cumyl-PEGACLONE, the only psychoactive substance detected despite relatively low concentrations in peripheral blood (0.09 ng/mL), was considered likely (TSS = 3). The death was caused by asphyxia due to SC-induced vomiting and aspiration of partially digested gastric content.

As shown in the present case series, the effects of 5F-Cumyl-PEGACLONE are consistent with vomiting, convulsions, loss of consciousness, and respiratory depression. The compound could additionally cause a decrease in body temperature and kidney injuries. The occurrence of four cases of death involving 5F-Cumyl-PEGACLONE in a relatively short period of time (October 2018–February 2019) is a matter of public health concern. The ‘relative safety’ of SCs bearing a γ-carbolinone core structure suggested in the context of the use of Cumyl-PEGACLONE [8] was not confirmed in our case series, possibly due to the 5-fluorination accounting for higher potency and potential higher toxicity, which are still to be precisely assessed.

Regarding postmortem redistribution, given the uncertainties regarding the pharmacokinetic profile of the substance, the time between 5F-Cumyl-PEGACLONE consumption and death, mechanisms of death, effects of cardiopulmonary resuscitation, and type of transport of the corpses, it is only possible to draw hypotheses. In case 1, it could be hypothesized an incomplete distribution of the substance at the moment of death due to very recent intake. In case 2, despite a very long PMI of 7 days, no difference between peripheral and central blood level was seen. Central to peripheral ratio (< 1, > 1 or = 1) has been historically used to assess postmortem redistribution, which in this case would be ruled out [40, 67]. However, in cases 3 and 4, which were characterized by a shorter PMI, concentrations were higher in heart blood than in femoral blood. Particularly in case 4, the central to peripheral ratio was 3.8. According to circumstantial data, the male had smoked approximately 7 h before being found dead, thus the SC absorption and distribution phase was probably over. It can be hypothesized that an initial postmortem redistribution may occur from highly concentrated organs (e.g. liver, lungs) to the central compartment in the first days after death, with a substantial equilibration of peripheral and central concentrations after a longer time interval. However, in-depth studies and further case reports/series are necessary to achieve a better understanding of this phenomenon, which has been described as a ‘toxicological nightmare’ [68].

From our data follows that for SCs, or at least for 5F-Cumyl-PEGACLONE, it seems not prudent to rely on the results of toxicological analyses performed with methods validated for serum when analyzing post-mortem material. However, if the available quantity of sample material does not allow standard addition for quantitation, this might be the only option and concentrations should be taken only as rough approximations then (in our cases there was a factor of 2 to 5 between the measured concentrations).

In the case series here presented, postmortem blood concentrations overlapped despite different TSS classification. This is a further confirmation that despite deaths involving SCs have increasingly been reported, the assessment of the toxicological significance is challenging and cannot rely on the mere results of toxicological analyses.

5F-Cumyl-PEGACLONE is an emerging and extremely potent SC, which raises serious public health concerns. Several issues concerning postmortem redistribution, toxic concentration ranges, and possible mechanisms of death remain unsolved and demand increased research efforts and further studies. Our results emphasize that toxicological results should be taken with caution when methods are applied to postmortem samples if they were validated for other matrices only and standard addition should be preferred.

In forensic toxicological postmortem case work, and especially when novel compounds are involved, a comprehensive analysis of circumstantial, clinical, and postmortem findings, as well as an in-depth toxicological analysis is indispensable for a valid interpretation.