Endocrine Abstracts

Introduction: Steroid profiling by mass spectrometry approaches consists in the simultaneous measurement of several steroid molecules in a biological sample, allowing an optimal characterization of steroidogenesis alterations, particularly in the context of adrenal tumors. Twenty-four hours urine samples have the advantage of being non-invasive and of giving an integrated view of steroidogenesis. Urinary steroid profiling has thus been shown to be particularly useful in the diagnosis of adrenal unilateral tumor (Weykamp, 1989; Arlt, 2011; Honour, 2018). We describe here the optimization and validation of a gas chromatography-mass spectrometry (GC-MS) approach, allowing the determination of a profile of 19 steroid urinary metabolites.

Material and methods: Urinary samples preparation required several steps including enzymatic hydrolysis, liquid-liquid extraction and derivatization of steroid metabolites. Retention time and mass spectrum of each steroid metabolite were determined by injection of the corresponding external standard in fullscan mode. Calibration curves were obtained by the sequential injection of a growing amount of each steroid external standard from 0.05 to 5 μg. A fixed amount of 8 internal standards was added to each sample and calibration point to normalize the results. Linearity, repeatability and reproducibility of the method were further evaluated.

Results: Optimized conditions of urine samples (U) preparation were as follows: enzymatic hydrolysis with arylsulfatase/glucuronidase for 4 hours at 55°, first derivation in methoxyamine/pyridine for 4 hours at 55° and second derivation in trimethylsilyimidazole for 5 hours at 100°. R² of the 19 calibration curves ranged from 0.983 to 0.998. Coefficient of variation (CV) of repeatability and reproducibility ranged respectively from 2.4% to 14.4% and from 3.2% to 13.6%. As previously described (Arlt, 2011), we confirmed with this approach, the increase in steroid precursor metabolites (including THS, 5-PT and 5-PD) in urinary samples from patients with adrenocortical carcinoma. This method was applied for the first time to urinary samples from 5 patients with Cushing disease and 10 control subjects matched on age and sex, highlighting a global activation of steroidogenesis in Cushing disease, characterized by an increase of both glucocorticoid (THF, ALLO-THF, THE), androgen (Etio, An), glucocorticoid and androgen precursors (PT, 5-PT, 5-PD) urinary metabolites.

Conclusion: We propose here a complete methodology of urinary steroid profiling by GC-MS, validated on samples from patients with adrenal carcinoma. This approach will give new insights into the characterization of steroidogenesis alterations, including in Cushing disease.