Development of a systematic workflow for early diagnosis and characterization of drug metabolism

Code: 21A05XT

The design and synthesis of new anabolic androgenic steroids, characterized by reduced side effects and better oral bioavailability, begun shortly after the structure of testosterone was identified in 1935. Since then, the discovery of new steroids has been accompanied by studies on their metabolism, to understand how they could be metabolized by phase I and II enzyme isoforms that alter their toxicity, action, and elimination. The presence of pharmacologically active metabolites, their variable formation, and their potential interaction with other drugs and/or nutritional supplements, may indeed interfere with the analytical results of doping controls. Furthermore, although metabolite detection allows identification of the metabolic pathways involved in drug absorption, molecules with related structures and similar physicochemical properties may follow generally common metabolic pathways, but different biotransformation steps, making more challenging the selective and unambiguous identification of the original compound. The process of detecting, characterizing, and confirming metabolites is time-consuming and resource-intensive. The analytical strategy currently used to identify drug metabolites involves several steps, some of which are more influenced by knowledge bias. The combination of knowledge-based and machine-learning approaches represents an innovative and effective way to predict, identify and characterize drug metabolites, not only in anti-doping but also in clinical or forensic fields. 

To this end, the proposed systematic analytical protocol, capable of combining targeted and untargeted analyses, full scan, MS/MS experiments, and previous knowledge, represents a strategic tool for the rapid detection of drug metabolites. The use of accurate mass measurements by high-resolution mass spectrometry (HRMS) and the complementarity of liquid and gas chromatography ensures a high degree of completeness to the results. The processing of the raw data by multivariate analysis allows to focus on the specific features of the condition, without performing any a priori bias.