Steroid profiling by gas chromatography/mass spectromy (GC/MS) and liquid chromatography/tandem mass spectrometry (LC/MS/MS) in patients with 46-XY DSD.
In a significant number of patients with 46,XY DSD the underlying cause is unknown. Steroid profiling by GC/MS and liquid chromatography/mass spectrometry (LC/MS) is an established diagnostic tool also in patients with single enzyme steroidogenic disorders causing 46,XY or 46,XX DSD. The advantage of GC/MS and LC/MS is the ability to visualise the complete steroid metabolome in a single patient based on the analysis of 24-h urine and plasma samples. Each steroidogenic disorder has a characteristic metabolic steroid “fingerprint” that can be readily detected. In this workpackage we will employ GC/MS and LC/MS/MS to discover novel steroidogenic disorders.
The WP will employ GC/MS and LC/MS/MS to analyse urine and plasma samples from the Pan-European 46,XY cohort provided by the members of the consortium. The main focus will be on prospectively collected urine (Birmingham) and plasma samples (Kiel) from all newly diagnosed 46,XY DSD patients. Prospective sample collection is carried out in the neonatal period and at 1, 3, 6, 12 months of age and then in annual intervals; subsequent analysis will be performed in Birmingham (GC/MS) and Kiel (LC/MS/MS), employing the state-of-the-art equipment available in the both sites.
The steroid output data will be entered into a DSD GC/MS database (Birmingham) and a DSD LC/MS/MS database (Kiel) that will be virtually linked to the Clinical DSD Registry (Glasgow) via the web-based database tool (WP01).
The steroid profiling data will be analysed by conventional analysis with quantitative description and plotting of data (Birmingham, Kiel). In addition, calculation of steroid substrate/product ratios for steroidogenic key enzymes will be implemented as a measure of specific enzymatic activity (Birmingham/Kiel). This allows via comparison to sex- and age-specific reference cohorts established in Birmingham and Kiel to detect evidence of enzymatic blocks. Furthermore, data analysis will be extended to unsupervised and supervised machine learning methods including principal component analysis (collaboration with Prof Manfred Opper, Professor of Artificial Intelligence, Technical University Berlin, Germany, and Prof Michael Biehl, University of Groningen, The Netherlands).
The GC/MS and LC/MS/MS steroid output databases will be virtually linked with the clinical database (WP01) to facilitate correlation of the steroid profiling results with the patient phenotype. WP05 will closely collaborate with WP02 (Gene Discovery) in a bidirectional manner, i.e. patients with novel mutations or rearrangements identified by WP02 will be analysed regarding their steroidogenic phenotype by WP05 and similarly patients identified within WP05 found to suffer from a specific steroidogenic disorder will be verified by WP02 via candidate gene sequencing, and patient subgroups identified via the computational approach in WP05 will undergo targeted analysis by comparative genomic hybridization within WP02.