Urine and plasma samples collected from three individuals suspected of HHC consumption were submitted to systematic forensic toxicological analysis that involved (a) immunological testing of urine samples , (b) GC–MS analysis of plasma samples targeting THC, 11-OH-THC and THC-COOH and (c) non-targeted LC–MS-MS analysis of urine as well as plasma samples. The obtained analytical results provided evidence that all three persons had consumed HHC.

Both urine samples tested positive using the CEDIA THC immunoassay, which primarily targets THC-COOH. Another plausible explanation for the observed results is the cross-reactivity of HHC metabolites. This hypothesis finds support in studies conducted in the 1980s, which highlighted that oxidized forms of HHC might lead to false-positive test results in immunoassays developed for detecting THC-COOH (16).

The applied GC–MS method utilized selected ion monitoring to target specific fragments of THC, 11-OH-THC and THC-COOH. The analysis of the three plasma samples yielded negative results for THC and 11-OH-THC. However, the results for THC-COOH were indeterminate. While a detectable peak was observed for each ion within the specified retention time window, the ion ratios deviated from the acceptable ranges. The occurrence of interfering ions related to one or more THC-related compounds (i.e., HHC metabolites) likely contributed to these deviations, though their exact origin remains speculative.

For non-target analysis, LC–MS-MS with data-dependent acquisition control was employed (13). Compound annotation was accomplished through tandem mass spectral library search. All samples tested positive for THC-COOH. This observation was indicative for a recent THC consumption. However, negative tests for THC, 11-OH-THC and other common psychoactive substances suggested that the observed impairment was linked to another compound that was not covered by the applied annotation workflow. To track down the identity of this substance, tentative matches to cannabigerol that were observed in the plasma samples proved useful. Cannabigerol is typically a minor constituent of cannabis, making it challenging to detect in plasma samples. To confirm the compound’s identity, a cannabigerol reference standard was analyzed under identical experimental conditions. The obtained mass spectrometric information, including the m/z of the precursor ion and the corresponding fragment ion spectrum, supported a putative match of the compound to cannabigerol (Figure 2a and b). However, due to the observed non-conformance in retention times, identity was excluded.

To find an alternative candidate, a PubChem search was conducted using the molecular formula of cannabigerol (C₂₁H₃₂O₂). Among the 3,779 results, one was identified as HHC. By analyzing an HHC reference standard, the identity of the unknown compound was verified as HHC. In this case, the mass spectrometric data (Figure 2c) as well as the chromatographic information perfectly matched.

HHC and cannabigerol are structurally related compounds, with cannabigerol being considered the open-chain form and HHC the related cyclic form. Due to ring-opening reactions involved in the tandem mass spectral fragmentation of cyclic cannabinoids, the fragment ion mass spectra of these compounds often exhibit a high degree of similarity to the spectra of their open forms (e.g., THC vs. CBD). This similarity also applies to HHC and cannabigerol, resulting in fragmentation patterns that are difficult to differentiate from each other (Figure 2). As a consequence, if a software-supported tandem mass spectral library search is utilized, HHC spectra may be matched to cannabigerol if the library contains cannabigerol-specific spectra but lacks HHC-specific spectra.

As previously mentioned HHC was detected only in the analyzed plasma samples (Figure 3a), suggesting that it may have undergone metabolic modifications before being excreted in urine. Drawing parallels with the human metabolism of THC and CBD, it was anticipated that metabolic transformation reactions would involve Position 11 as well as the pentyl side chain (6–8). Additionally, it was expected that both the parent drug and its oxidized forms might undergo glucuronidation reactions (9). Leveraging the available LC–MS-MS data, these hypotheses were verified, enabling insights into the biotransformation pathways of HHC in humans.

In a first attempt to identify the expected HHC metabolites, we calculated the m/z values of protonated molecular ions and constructed corresponding XICs (Figure 3). By utilizing accurate mass information, we were able to provide evidence for the occurrence of oxidation reactions. These reactions included (i) hydroxylations (+O; Δm/z = 15.9949; Figure 3c), (ii) methyl oxidation to carboxylic acid (+O₂, −H₂; Δm/z = 29.9741; Figure 3e), and (iii) pentyl side chain oxidation to carboxylic acid (−(CH₂)₂, +O₂, (−H₂; Δm/z = 1.9429; Figure 3g). Furthermore, we observed signs of glucuronidation for both HHC and its Phase I metabolites (+C₆H₈O₆; Δm/z = 176.0321; Figure 3b, d, f and h).

Remarkably, two forms of putative HHC glucuronide were detected (M1.1 and M1.2; Figure 3b). Glucuronidation of HHC is limited to the hydroxyl group located at C1. As semi-synthetic HHC consists of a mixture of (9S)- and (9R)-HHC epimers, the glucuronidation process yields diastereomers that can be chromatographically separated (Figure 3b).

Furthermore, multiple peaks were observed in the XIC traces specific for hydroxylated HHC forms and the corresponding glucuronides (Figure 4c and d). The presence of different sites of modification and the formation of diastereomers offered plausible explanations for the occurrence of multiple peaks in the glucuronide chromatograms.

Crucially, some of the detected hydroxylated forms (M3.1, M3.2, M4.1 and M4.2; Figure 3c) were considered to be interferences only. These interferences were attributed to in-source fragmentation of the corresponding glucuronides (Figure 3d). In-source fragmentation of glucuronides generates ions identical to those of the precursors (17), which are then detected in MS mode and contribute to signals in the precursors’ XIC traces at the retention times of the glucuronides. As a result, among the five hydroxylated HHC species detected, only one was considered to be present in the unconjugated form (M2; Figure 3c), while the other four species were identified as glucuronides (M3.1, M3.2, M4.1 and M4.2, Figure 3d).

In the XIC traces of the carboxylic acids as well as of the corresponding glucuronides, only one prominent peak was observed (M5, M6, M7 and M8; Figure 3e–h). This suggests that either no diastereomers were formed in these cases, or if they were formed, the chromatographic method used was not capable of separating the diastereomers. Minor peaks in the XIC traces of the carboxylic acids at the retention times of the carboxylic acid glucuronides indicated in-source fragmentation of the glucuronides (Figure 3e–h).

To provide additional evidence for the proposed structures of HHC metabolites, we examined the available tandem mass spectral data. Spectra representing the individual metabolites are provided in msp-format in the electronic supplementary material.

Figure 5 displays the tandem mass spectra of several glucuronides, specifically selected as representative examples due to their ability to provide structural information on both the conjugated and corresponding unconjugated forms (18). The dissociation of the glucuronides involved neutral loss of the glucuronic acid (Δm/z = 176.0321) and the generation of fragment ions identical to those of the precursor ions of the original forms (Figure 5). Certain secondary fragment ions were crucial for identifying the site of modification.

An important fragmentation reaction shared by HHC and other cannabinoids involves cleavage of bonds between Positions 10 and 10a, 5 and 6 as well as 6a and 10a (Figure 1b). The resulting (2,6-dihydroxy-4-pentylphenyl) methylium ion retains the pentyl side chain while losing the Positions 8–11. In the case of HHC, this specific fragment ion exhibits an m/z value of 193.1223 (Figure 2c), which was indeed observed in the fragment ion mass spectra of the putative HHC glucuronides (M1.1 and M1.2; Figure 4a). This fragmentation reaction proved valuable for distinguishing hydroxylation at Positions 8–11 from hydroxylation at the pentyl side chain (Figure 1a). For the former species (M3.1 and M3.2), the fragment ion at m/z 193.1223 was detected (Figure 4b). For the latter species (M4.1 and M4.2), two fragment ions with specific m/z shifts to 209.1159 (+O; Δm/z = 15.9918) and 191.1082 (+O, −H₂O; Δm/z = −2.0152) were observed (Figure 4c).

Carboxylic acids, including THC-COOH, are known to exhibit characteristic neutral losses of −H₂O (Δm/z = 18.0106) and subsequent −CO (Δm/z = 46.0055) during collision-induced dissociation (19). The accurate mass information provided evidence of two HHC-derived carboxylic acid species as well as their corresponding glucuronides, with oxidation reactions occurring at C11 (M5 and M6) and C3′ (M7 and M8). In collision-induced dissociation, cleavage of water and CO was observed for both putative HHC metabolites (Figure 4d and e). Nevertheless, with the available fragmentation information, including reference data from THC-COOH, unequivocal verification of the chemical structures was only achieved for 11-nor-9-carboxy-hexahydrocannabinol (HHC-COOH, M5) and its glucuronide (M6). For the other species, doubts about the correctness of the proposed structures still exist. Further analyses, involving specific in vitro assays and chemical synthesis (5, 7, 20), would be required to increase identification confidence and to verify the preference for C3′ over other positions of the pentyl side chain for oxidation.

In an attempt to validate accomplished annotations, we compared the tandem mass spectral information of selected metabolites with data obtained from analyzing reference standards. The reference standards analyzed included 8-hydroxy-HHC, 9-hydroxy-HHC, 10-hydroxy-HHC, 11-hydroxy-HHC as well as 11-nor-9-carboxy-HHC. The results are provided in the electronic supplementary material Figures S1 and S2. The different hydroxylated reference compounds exhibited very similar fragmentation patterns, and the observed fragment ions corresponded well with those observed for related compounds in the human samples (M2, M3.1 and M3.2). Additionally, the fragment ion mass spectrum of the 11-nor-9-carboxy-HHC standard aligned with the corresponding spectra annotated in the human samples (M5 and M6).

Based on the LC–MS-MS data, the study was able to identify 11 HHC metabolites, and an overview of the compounds detected in the analyzed plasma and urine samples is provided in Table I. The proposed metabolic transformation reactions are summarized in Figure 5. Transformation of HHC involves the formation of HHC glucuronide (M1), HHC-COOH (M5) and HHC-COOH glucuronide (M7). Hydroxylation appears to occur at either Positions C8 to C11 or the pentyl side chain. Given the observation of the corresponding carboxylic acid, it is highly likely that C11 is the primary site of oxidation (M2), followed by glucuronidation (M3.1 and M3.2). While confirmation is still pending, there is some evidence supporting oxidation of C3′, resulting in hydroxylated and carboxylic acid species (M7), along with their corresponding glucuronides (M4.1, M4.2 and M8).