Legally purchasable cannabis containing less than 1% THC but, at the same time, relevant levels of CBD has become a popular product for self-medication and non-medical recreational use [26]. Although research has mainly focused on blood levels and their impact on psychomotor functions following the consumption of illegal cannabis products [44,45,46], studies on the pharmacokinetics and psychopharmacology of low-THC/high-CBD cannabis consumption have gained more attention over the years [19, 26, 47, 48]. In the forensic context, so far, two psychopharmacological studies from Switzerland [19] and Italy [26] address the serial THC and CBD blood concentrations required to determine the pharmacokinetic profiles of controlled low-THC/high-CBD cannabis smoking or inhalation and its effects on adults’ vigilance, cognitive, and motor skills using different laboratory tests. However, both studies observed high inter-individual variability in THC or CBD blood concentrations that could not be attributed to gender or age. Instead, it suggests that the result is due to the participants’ different smoking habits (e.g., some subjects may have held the smoke in their lungs for a longer time) or smoking times (e.g., 6–23 min, study arm 1 (n = 27), product 1 [19]). Taking into account differences in the products used, demographics (e.g., size, BMI), and smoking habits (e.g., consumption patterns), the blood concentration results of THC, its metabolites, and CBD were in line with previous findings. Nevertheless, to date, little is known about the impact of changes in activity of cannabinoid-metabolizing enzymes on plasma or whole blood concentrations of CBD, THC, and their metabolites [27].

Of interest in the context of THC metabolism are CYP2C9 and CYP2C19, as they were shown to contribute to THC phase I metabolism [29]. Importantly, numerous in vitro and in vivo studies have shown the impact of frequently occurring genetic variants on the activity for these enzymes, and consequently, on the pharmacokinetics of their substrates [49,50,51]. In the context of THC metabolism, reduced enzymatic activity and changes in the conversion of THC to THC-OH have been demonstrated in vitro for the CYP2C9*2 and CYP2C9*3 variants, showing up to a 70% reduction in intrinsic clearance [52, 53]. Limited in vivo data further support a role of CYP2C9 showing an impact of the low-function allele (CYP2C9*2) and the non-functional allele CYP2C9*3 on the metabolic ratio (THC/THC-COOH) [27]. In pharmacological studies, these alleles are commonly used to predict the respective CYP2C9 phenotype of an individual, applying defined activity scores [54].

Even if the findings by Gasse et al. [27] support a role of CYP2C9 in cannabinoid metabolism, there is currently no data available on whether the genotype-predicted CYP2C9 phenotype would also reflect this observation. Moreover, there is currently no information on whether the genotype predicted CYP2C19 phenotype is of relevance. In this study, we selected the most frequent genetic variants of CYP2C9 and CYP2C19 and determined them in leftover samples from a previous study where research data on blood concentrations were available [19]. We subsequently predicted the respective phenotypes to perform a post-hoc analysis, while being aware of the low number of eligible individuals and, consequently, the low number of poor metabolizer phenotypes (e.g., CYP2C19 2.4%, and CYP2C9 2.6%) to be expected. Indeed, we only found normal (NM, 56%) and intermediate (IM, 44%), but no poor metabolizers. The analyzed samples, originally collected in Switzerland [19], did not include information on the biogeographic origin of the participants. Nevertheless, we obtained phenotype frequencies comparable to those of the European population (62.8% NM, 34.4% IM [41]),. Similarly, for CYP2C19, we observed 48.1% NM, 25.9% IM, 22.2% RM and 3.7% UM, aligning with the frequencies present in the European population (39.6% NM, 26.3% IM, 27.1% RM and 4.6% UM [42]),, and again without a poor metabolizer (2.4% [42]),.

We restricted our analysis to CYP2C9 and CYP2C19, even though CYP3A4 is known to contribute to the metabolism of THC and CBD [29, 30, 35, 53]. However, the variability of CYP3A4 activity depends less on genetic predisposition and is mainly affected by non-genetic factors such as drug-drug interactions in which phenomena like induction and inhibition are known to affect an individual’s CYP3A4 phenotype greatly [55]. In addition, non-genetic factors (e.g., concomitant medication, diet) [56, 57] and CYP enzyme inhibition due to cannabis constituents themselves (e.g., CBD) [58] may also affect the genotype-predicted phenotypes, which is also true for CYP2C9 and CYP2C19 investigated in this study. Even if long-term medication, and no prior illegal substance or alcohol abuse was an exclusion criterion for the original study [19], the phenomenon of phenoconversion needs further consideration [59]. It could be addressed in a future prospective study on THC and CBD by the application of a phenotyping cocktail as previously proposed [60, 61].

Assessing the influence of the genotype-predicted phenotypes of CYP2C9 on blood concentrations of THC, its primary metabolites (THC-OH and THC-COOH), and CBD [19] revealed no significant differences between CYP2C9 NMs and IMs in the single consumption group. This finding was further supported by comparing the AUC₀min–END for THC, its metabolites, and CBD between the two predicted phenotype groups. Furthermore, metabolic ratios were compared to evaluate phase I metabolism, showing no significant differences between CYP2C9 NMs and IMs. Although a slower metabolism of THC might be expected in the IM group due to reduced CYP2C9 enzymatic activity [53, 62], this anticipated effect was not observed following a single inhalation. Given the controlled setting and route of administration in the study [19], we hypothesize that during single inhalative dosing, the influence of CYP2C9 phenotypes may be minimal, potentially due to a metabolic shift from CYP2C9-dependent pathways to alternative routes, such as those mediated by CYP3A4 [29, 53].

Analyzing the NM and IM CYP2C9 phenotypes on blood concentrations of THC, THC-OH, THC-COOH, and CBD in the repeated consumption group, the THC-COOH levels were slightly lower in the IM group, aligning with findings from previous studies [27, 62], which reported reduced THC-COOH concentrations in individuals with the CYP2C9 reduced function alleles compared to those with the CYP2C9*1/*1 genotype. While the above-mentioned studies [27, 62] noted similar to higher THC levels in function impairing genotypes compared to the CYP2C9*1/*1 genotype, we observed a trend toward lower THC concentrations in IMs compared to NMs, with statistically significant differences at time points T = 60 min and T = 180 min. Moreover, earlier studies showed similar THC-OH levels between the genotype groups. In contrast, for the IM group in this research, the THC-OH appeared higher shortly after inhalation but converged with NM levels at later time points. Interestingly, these differences were not observed in the single consumption group, suggesting that repeated cannabis use may induce metabolic activity more prominently in IMs than in NMs. Specifically, this may enhance the CYP2C9-mediated hydroxylation of THC to THC-OH, while the subsequent conversion of THC-OH to THC-COOH remains comparatively slower in IMs. This pattern, along with the significantly lower THC blood levels in IMs than in NMs, supports the hypothesis that THC is more rapidly converted to THC-OH in IMs, which may then accumulate due to a reduced rate of further metabolism to THC-COOH.

Concerning CBD, significantly lower blood levels were observed for IMs compared to NMs at the time points T = 10 min, T = 20 min, and T = 60 min, as observed for the THC metabolism. However, due to the lack of data on the metabolites of CBD, no further evaluation is possible. Nevertheless, the analysis of the AUC₀min–END values and the previously described metabolic ratios for THC and its metabolites yielded comparable results between the two phenotype groups. Given the small sample size (n = 4) and the resulting limited statistical power [63], the observed differences in THC, THC-OH, and CBD blood concentrations should be interpreted with caution, as they may be overestimated.

When examining the influence of the CYP2C19 phenotypes in the single consumption group, again no significant differences in THC, THC-OH, THC-COOH, or CBD blood levels, AUC, or metabolic ratios were observed between phenotypes of NM, IM, RM and UM, consistent with previous findings on the effect of CYP2C19 genotypes on THC whole blood levels [27]. Although statistical analysis could not be performed for the repeated consumption group due to the small sample size, descriptive data indicate no meaningful differences in blood concentrations of any compound across the various CYP2C19 phenotypes. This outcome is consistent with the relatively minor role of CYP2C19 in the hydroxylation of THC, its activity accounting for only about 1% of that of CYP2C9 [29]. Based on our results and the limited contribution of CYP2C19, we conclude that variations in its activity are unlikely to have a measurable impact on the metabolism of either THC or CBD. Therefore, at first glance, CYP2C19 phenotyping does not appear to be relevant for the interpretation of toxicological findings in DUID cases of cannabis consumption.

Although our findings exhibit trends similar to those reported in previous studies (lower THC-COOH levels in the IM group), the observed effects appear to be less pronounced. One possible explanation lies in the route of administration: in earlier studies, THC was either administered orally [62] or the route of intake was not clearly documented [27]. Oral ingestion is subject to a significant first-pass effect in the liver, which strongly influences the metabolism of THC and CBD [64, 65], potentially amplifying genotype-related differences. In contrast, our study employed a controlled inhalation method, which bypasses the first-pass metabolism and may therefore reduce the observable impact of genetic variation. However, a key limitation of the presented study is the small sample size and variability in pulmonary uptake due to inhalation technique, a constraint inherent to its retrospective design, and a lack of poor CYP2C9 metabolizers. Particularly, those are relevant since Sachse-Seeboth et al. [62] reported the most pronounced metabolic differences in individuals with the CYP2C9 *3/*3 genotype. Including them would likely enable a more comprehensive assessment of the potential influence of genotype-predicted metabolism on cannabinoid pharmacokinetics, providing a better understanding of the relevance of pharmacogenetic variability in this field. In the long term, this could be of great importance to forensic examiners in the interpretation of cannabis-related DUID cases, as it would provide further relevant information on the complex field of cannabis metabolism and help to understand the mechanism of inter-individual variability in blood concentrations and effects of cannabis users.

Finally, as mentioned, cannabinoid metabolism is known to be complex, characterized by multiphase pharmacokinetics and long terminal half-lives. Thus, the evaluation of blood concentrations is non-trivial and is also influenced by rediffusion from adipose and other poorly perfused tissues into blood. Furthermore, excretion is considered to be altered due to an enterohepatic recirculation of phase II metabolites (e.g., glucuronidated THC-COOH) [66], indicating that biliary excretion as route of elimination for cannabinoid conjugates may be an important factor to consider when analyzing blood elimination profiles of cannabinoids [67]. However, investigation of phase-II metabolites would require a different sample preparation (e.g., pre-treatment with glucuronidase) or instrumental analytics (e.g., LC-MS) than those applied in the original study. Here, the method comprised automated on-line solid-phase extraction (SPE) and derivatization [19, 32] followed by detection using GC-MS/MS, which detects only the free (non-glucuronidated) fraction of analytes. Thus, in this analysis, only phase-I metabolites were considered, even though UGT polymorphism and its impact on UGT activity could be of additional interest for further studies [31, 68].