Hye Ju Yu (한국과학기술연구원)

Man Ho Choi (한국과학기술연구원)

Hye Ju Yu (한국과학기술연구원)

Yinglan Xu (한국과학기술연구원)

Moonseok Choi (한국과학기술연구원)

Man Ho Choi (한국과학기술연구원)

Anabolic androgenic steroids (AASs) are synthetic variations of androgens and their prohormones, which are the most frequently misused substances in doping control. In addition to parent compound, it metabolites can provide the abuse of the corresponding AAS, and exploring metabolite identification and gaining comprehensive metabolite information are imperative. In this study, a GC–Orbitrap–MS with full-scan acquisition enables retrospective non-targeted data analysis. An in-house database was constructed with 153 synthetic and endogenous steroids, as their trimethylsilyl derivatives. The library contains accurate mass, retention index (RI), full-scan spectra, characteristic fragment ions, and structural information for each compound. Target screening was performed using five identification criteria: mass accuracy (< 5 ppm), %ΔRI (< 1%), isotopic pattern score (> 70), diagnostic fragment ion matching (at least three ions from the in-house database), and full-scan spectral comparison (similarity score > 30 against the spectral library). Initially, seven metabolites in AASs, including methyltestosterone, oxandrolone, and stanozolol, were successfully identified at concentrations ranging from 50% to 400% of the WADA Minimum Required Performance Level. By employing the RI, the method offers improved robustness against variations in chromatographic conditions, such as oven temperature gradients or column differences, thereby facilitating potential inter-laboratory database sharing. The established workflow is expected to provide comprehensive steroid data and detection windows, contributing to enhanced the sensitivity and accuracy of anti-doping assays.