Delineating Pediatric Adrenocortical Tumors by GC-MS Urinary Steroid Metabolome Analysis: Observations From the MET Study

Abstract Introduction Adrenocortical tumors (ACTs) comprise adrenocortical adenomas (ACAs) and adrenocortical carcinomas (ACCs); the latter are highly aggressive. Pediatric adrenocortical tumors (pACTs) are functional and thus symptomatic. We investigated whether gas chromatography-mass spectrometry (GC-MS) urinary steroid metabolome analysis can detect tumors and differentiate between adenomas and carcinomas. Method We investigated 46 patients with pACTs (age: median 6.9; range 0.7-17 years; 36 females; 21 ACAs, 25 ACCs) from the German Society for Pediatric Oncology and Hematology for Malignant Endocrine Tumors study at the time of initial diagnosis. Thirty-six urinary C21-, C19-, and C18-steroid metabolites were quantified from spot urines by targeted GC-MS. Data underwent computational analysis by log transformation followed by z-transformation and machine learning classifiers, such as cluster and principal component analysis, logistic regression, and decision trees. Results pACTs and controls showed highly significant differences (P < .001) for most parameters, particularly gross elevation of the urinary metabolites of the Δ5-steroids pregnenolone, 17-hydroxypregnenolone, and dehydroepiandrosterone, as well as of tetrahydro-11-deoxycortisol (THS), the urinary main metabolite of 11-deoxycortisol. Pediatric adrenocortical carcinomas (pACCs) showed higher THS and decreased corticosterone metabolites (P < .05). Selected biomarkers discriminated pACTs from controls, as well as pACCs from pediatric adrenocortical adenomas (pACAs) with high sensitivity and specificity. Conclusion Urinary GC-MS steroid metabolome analysis is noninvasive and allows good detection of ACTs from spot urine as well as good differentiation of pACCs from pACAs. The dominance of Δ5-steroids in pACTs points to the zona reticularis as a common site of origin. The parameters differentiating pACAs from pACCs differ from adults.