ESPE Abstracts

Introduction: ACTs comprise adenomas (ACAs) and carcinomas (ACCs), the latter having a poor prognosis. In children, ACTs are functional and thus symptomatic. We investigated whether GC-MS urinary steroid metabotyping can detect tumors and differentiate between adenomas and carcinomas.

Method: We investigated 46 patients (median 6.9; range 0.7-17 yrs; 36 females) with ACTs from the GPOH-MET Study (Registry of the German Society for Pediatric Oncology and Hematology for Malignant Endocrine Tumors) at the time of initial diagnosis. Patients were recruited between 2001 and 2024. n = 21 were diagnosed with ACAs and n = 25 with ACCs according to histopathological criteria. n = 145 urines from healthy children served as controls. 36 Steroid metabolites were quantified (µg/l) from spot urines by targeted GC-MS urinary steroid metabolome analysis. Relative enzyme activities were calculated according to typical precursor / product metabolite ratios. Log₂ and z-transformed metabolite data underwent supervised classification analysis by univariate logistic regression followed by ROC-Analysis in order to find most discriminating metabolites. In addition machine learning classifiers were applied (results not shown here).

Results: ACTs differed (OR 4.62-7.50, P <0.001) from controls primarily in pronounced elevations of the metabolites of all 5-ene-steroids (i.e. pregnenolone, 17-hydroxypregnenolone and DHEA), decreased activity of 3β-hydroxysteroiddehydrogenase (3βHSD), gross elevation of THS (a tetrahydrogenated metabolite of 11-deoxycortisol) and reduced 11-hydroxylase activity. Selected biomarkers discriminated ACTs from controls by an accuracy > 0.81, respectively (ROC Analysis). ACCs differed (OR 2.03-2.68, P <0.001) from ACAs primarily in decreased corticosterone metabolites, increased 17-hydroxylase activity and reduced 5α-reductase activity. These biomarkers discriminated ACCs from ACAs by an accuracy > 0.7, respectively (ROC Analysis).

Conclusion: 1) Urinary targeted GC-MS steroid metabotyping is noninvasive and allows good discrimination between ACTs and controls from spot urine. 2) In ACTs, the dominance of 5-ene-steroids matches with reduced 3βHSD activity and points to the zona reticularis as a common site of origin for ACCs and ACAs. 3) Pediatric ACAs and ACCs differ mainly in their activities of 17-hydroxylase and 5α-reductase, a finding different from adults. 4) Our findings provide evidence that pediatric ACTs behave differently from those in adults. Thus, findings in adult ACT patients cannot be transferred to pediatric patients.