Metabotypes of congenital adrenal hyperplasia in infants determined by gas chromatography-mass spectrometry in spot urine.

Biochemical monitoring of treatment in infants with classic congenital adrenal hyperplasia (CAH) is not yet well defined. The aim of this study was to perform a cluster analysis of the urinary steroid metabolome for treatment monitoring of infants with classic salt-wasting CAH. We analyzed spot urine samples obtained from 60 young children ≤ 4 years of age (29 females) with classic CAH due to 21-hydroxylase deficiency treated with hydrocortisone and fludrocortisone by targeted gas chromatography-mass spectrometry (GC-MS). Patients were classified into different groups according to their metabolic patterns (metabotypes) using unsupervised k-means clustering algorithms. Three metabotypes could be discovered. Metabotype #1 (N = 15 (25%)) showed high concentrations of androgen and 17-hydroxyprogesterone (17OHP) precursor steroids, metabotype #2 (N = 28 (47%)) revealed balanced metabolic control, and metabotype #3 (N = 17; 28%) demonstrated severe adrenal suppression with low concentrations of androgen and 17OHP precursor steroids. Daily hydrocortisone doses and urinary concentrations of cortisol and cortisone metabolites did not differ between all three metabotypes. Metabotype #2 had highest daily dose of fludrocortisone (p = 0.006). Receiver operating characteristic curve analysis showed that 11-ketopregnanetriol (area under the curve [AUC] 0.967) and pregnanetriol (AUC 0.936) were most suitable of separating metabotype #1 from #2. For separation between metabotypes #2 vs. #3, the 11-oxygenated androgen metabolite 11-hydroxyandrosterone (AUC 0.983) and the ratio of 11-hydroxyandrosterone to tetrahydrocortisone (AUC 0.970) were most suitable. In conclusion, GC-MS-based urinary steroid metabotyping is a new method to help monitor the treatment of infants with CAH. This method allows classification of under-, over- and adequately treated young children.