In this study, we delineated the poorly characterized metabolism of anamorelin, a growth hormone secretagogue receptor agonist, in vitro using human liver microsomes (HLM), based on classical molecular networking (MN) and feature-based molecular networking (FBMN) from the Global Natural Products Social Molecular Networking platform. Following the in vitro HLM reaction, the MN analysis showed 11 neighboring nodes whose information propagated from the node corresponding to anamorelin. The FBMN analysis described the separation of six nodes that the MN analysis could not achieve. In addition, the similarity among neighboring nodes could be discerned via their respective metabolic pathways. Collectively, 18 metabolites (M1–M12) were successfully identified, suggesting that the metabolic pathways involved were demethylation, hydroxylation, dealkylation, desaturation, and N-oxidation, whereas 6 metabolites (M13a*-b*, M14a*-b*, and M15a*-b*) remained unidentified. Furthermore, the major metabolites detected in HLM, M1 and M7, were dissimilar from those observed in the CYP3A4 isozyme assay, which is recognized to be markedly inhibited by anamorelin. Specifically, M7, M8, and M9 were identified as the major metabolites in the CYP3A4 isozyme assay. Therefore, a thorough investigation of metabolism is imperative for future in vivo studies. These findings may offer prospective therapeutic opportunities for anamorelin.

Drug solubility limits intravenous dosing for poorly water-soluble medicines, which misrepresents their bioavailability estimation. The current study explored a method using a stable isotope tracer to assess the bioavailability of drugs that are poorly water-soluble. HGR4113 and its deuterated analog, HGR4113-d7, were tested as model drugs. To determine the level of HGR4113 and HGR4113-d7 in rat plasma, a bioanalytical method using LC-MS/MS was developed. The HGR4113-d7 was intravenously administered to rats that were orally pre-administered HGR4113 at different doses; subsequently, the plasma samples were collected. HGR4113 and HGR4113-d7 were simultaneously determined in the plasma samples, and bioavailability was calculated using plasma drug concentration values. The bioavailability of HGR4113 was 53.3% ± 19.5%, 56.9% ± 14.0%, and 67.8% ± 16.7% after oral dosages of 40, 80, and 160 mg/kg, respectively. By eliminating the differences in clearance between intravenous and oral dosages at different levels, acquired data showed that the current method reduced measurement errors in bioavailability when compared to the conventional approach. The present study suggests a prominent method for evaluating the bioavailability of drugs with poor aqueous solubility in preclinical studies.