The continuous emergence of new psychoactive substances (NPS), combined with varying analytical abilities and challenges in manage toxicological data, makes systematic toxicological analysis (STA) a persistent challenge. Forensic toxicology labs must implement comprehensive, up-to-date strategies that integrate different analytical and data analysis techniques, including retrospective analyses, to effectively detect and identify emerging substances. This study aimed to develop an indirect screening approach for monitoring New Psychoactive Substances (NPS), particularly new synthetic opioids, by employing LC-HRMS/MS analysis to evaluate variations in endogenous urinary metabolite profiles that reflect systemic responses triggered by their intake. The experimental design utilized in vivo CD-1 mice models, with twenty animals equally divided by sex (10 males, 10 females) receiving different treatments. On day 0, mice were dosed with a vehicle, and urine samples were collected during 0-12 h and 12-24 h intervals. On day 1, half of the animals from each sex received brorphine, while the other half were administered etonitazene; urine was again collected at the same time intervals. A second drug administration with subsequent urine collection occurred on day 8. Finally, following a one-week washout period, urine samples were collected once more on day 15. By combining multivariate and univariate analyses, most endogenous altered metabolites after drug administration were identified, including those related to lipid peroxidation, inflammation, and oxidative stress. ANOVA Simultaneous Component Analysis (ASCA) [1] was used to assess how different factors such as collection time, gender, and drug treatment affect metabolite variations. It decomposes total variation by experimental factors, applies PCA to simplify and interpret these effects, and uses permutation tests to confirm their statistical significance. This approach identifies which factors influence the data, highlights key variables, and aids in uncovering distinct metabolic patterns.
Smilde A. K., Jansen J.J. et al., bioinformatics, (2005), Vol. 21 no. 13, p. 3043–3048 doi:10.1093/bioinformatics/bti476