The very high frequency of prednisolone detection in the samples (78.5%), the low concentration of this steroid and, importantly, the narrow range of the 95% confidence limits (0.97-1.05 in MS(2) mode and 0.88-1.04 in MS(3) mode), could represent evidence that its presence is endogenous. In the light of these results, this hypothesis seems the most probable, even if further studies are required to confirm it. Furthermore, a microbiological origin (i.e. fermentation of cortisol after sample collection) could not be disregarded.

The possibility of an endogenous presence of the glucocorticoid prednisolone has already been demonstrated in bovine and horse urine, with the aim of clarifying its origin in this matrix, which is used by official agencies for the control of illicit treatments. From this point of view, the endogenous nature of prednisolone could be a major topic in doping control of both amateur and professional human athletes. A study was therefore made on 34 human volunteers (13 males and 21 females; aged 22-62) to detect the presence of prednisolone in their urine by HPLC-MS(3). One of the volunteers underwent vernal allergy treatment with betamethasone for two subsequent years. An investigation was carried out with the aim of verifying if the suppression, and the circadian rhythm, of cortisol urinary levels could also apply to prednisolone. The results of the study show that prednisolone was present in the urine of all 34 volunteers, with a concentration very close to 100-times lower that of cortisol, with no dependence on gender. The same ratio (1/100) was observed in the prednisolone and cortisol levels detected during the 24h together with the suppression of prednisolone by betamethasone treatment. These data demonstrate the endogenous nature of low concentrations of prednisolone in human urine, and motivate further studies about the biosynthetic pathways of this corticosteroid and its relationship with stress in humans, as already described in cows.

The objective of the current research was to develop a liquid chromatography-MSn (LC-MSn) methodology for the determination of free cortisol and its 15 endogenous metabolites (6β-hydroxycortisol, 20α-dihydrocortisol, 20α-dihydrocortisone, 20-β-dihydrocortisol, 20β-dihydrocortisone, prednisolone, cortisone, α-cortolone, β-cortolone, allotetrahydrocortisol, 5α-dihydrocortisol, tetrahydrocortisol, allotetrahydrocortisone, 5β-dihydrocortisol, tetrahydrocortisone) in human urine. Due to its optimal performance, a linear ion trap operating in ESI negative ion mode was chosen for the spectrometric analysis, performing MS3 and MS4 experiments. The method was validated for limit of detection (LOD) and limit of quantification (LOQ) (0.01 ng mL−1 and 0.05 ng mL−1, for all compounds, respectively), intra- and inter-day precision (CV = 1.4–9.2% and CV = 3.6–10.4%, respectively), intra- and inter-day accuracy (95–110%), extraction recovery (65–95%), linearity (R2 > 0.995), and matrix effect that was absent for all molecules. Additionally, for each compound, the percentage of glucuronated conjugates was estimated. The method was successfully applied to the urine (2 mL) of 50 healthy subjects (25 males, 25 females). It was also successfully employed on urine samples of two patients with Cushing syndrome and one with Addison’s disease. This analytical approach could be more appropriate than commonly used determination of urinary free cortisol collected in 24-h urine. The possibility of considering the differences and relationship between cortisol and its metabolites allows analytical problems related to quantitative analysis of cortisol alone to be overcome. Furthermore, the developed method has been demonstrated as efficient for antidoping control regarding the potential abuse of corticosteroids, which could interfere with the cortisol metabolism, due to negative feedback on the hypothalamus-hypophysis-adrenal axis. Lastly, this method was found to be suitable for the follow-up of prednisolone that was particularly important considering its pseudo-endogenous origin and correlation with cortisol metabolism. Graphical abstract