Comparison of in vitro and in vivo models for the elucidation of metabolic patterns of 7‐azaindole‐derived synthetic cannabinoids exemplified using cumyl‐5F‐P7AICA

Abstract Due to the dynamic market involving synthetic cannabinoids (SCs), the determination of analytical targets is challenging in clinical and forensic toxicology. SCs usually undergo extensive metabolism, and therefore their main metabolites must be identified for the detection in biological matrices, particularly in urine. Controlled human studies are usually not possible for ethical reasons; thus, alternative models must be used. The aim of this work was to predict the in vitro and in vivo metabolic patterns of 7‐azaindole‐derived SCs using 1‐(5‐fluoropentyl)‐ N ‐(2‐phenylpropan‐2‐yl)‐1 H ‐pyrollo[2,3‐ b ]pyridin‐3‐carboxamide (cumyl‐5F‐P7AICA) as an example. Different in vitro (pooled human liver S9 fraction, pooled human liver microsomes, and pig liver microsomes) and in vivo (rat and pig) systems were compared. Monooxygenase isoenzymes responsible for the most abundant phase I steps, namely oxidative defluorination (OF) followed by carboxylation, monohydroxylation, and ketone formation, were identified. In both in vivo models, OF/carboxylation and N ‐dealkylation/monohydroxylation/sulfation could be detected. Regarding pHS9 and pig urine, monohydroxylation/sulfation or glucuronidation was also abundant. Furthermore, the parent compound could still be detected in all models. Initial monooxygenase activity screening revealed the involvement of CYP2C19, CYP3A4, and CYP3A5. Therefore, in addition to the parent compound, the OF/carboxylated and monohydroxylated (and sulfated or glucuronidated) metabolites can be recommended as urinary targets. In comparison to literature, the pig model predicts best the human metabolic pattern of cumyl‐5F‐P7AICA. Furthermore, the pig model should be suitable to mirror the time‐dependent excretion pattern of parent compounds and metabolites.