Genotype-refined 17OHP cut-offs diagnosing nonclassical CAH due to 21OH deficiency in children with premature pubarche

Abstract Context Premature pubarche (PP) is a frequent reason for endocrine evaluation. Between 5% and 20% of children presenting with PP may have 21OHD nonclassic congenital adrenal hyperplasia (21OHD-NCAH). Objective To define optimal basal and ACTH-stimulated 17OHP cutoffs for diagnosing 21OHD-NCAH in children with PP, using CYP21A2 genotyping as the gold standard. Design, setting, and patients A diagnostic accuracy study including 203 children with PP (median age: 7.5 years [6.4-8.3]; 85% female) who underwent ACTH stimulation testing. Biallelic pathogenic variants upon CYP21A2 genotyping defined 21OHD-NCAH. Main outcome measure(s) Diagnostic performance of basal and post-ACTH 17OHP levels assessed by ROC curve analysis and thresholds defined according to the Youden index. Results 21OHD-NCAH was confirmed in 32 children (15.7%; 7.1 years [6.1-8.6]; 78% female). Basal 17OHP demonstrated excellent diagnostic accuracy (AUC = 0.98). A cutoff of 170 ng/dL (5.1 nmol/L) provided the best diagnostic balance, with 97% sensitivity and 91% specificity. A higher threshold of 410 ng/dL (12.4 nmol/L) achieved 100% specificity but reduced sensitivity (75%), whereas a lower cutoff of 118 ng/dL (3.5 nmol/L) yielded 100% sensitivity at the expense of specificity (84%). A post-ACTH 17OHP > 1104 ng/dL (33.4 nmol/L) yielded 100% sensitivity and specificity. Androgen levels overlapped substantially, and no clinical feature reliably distinguished 21OHD-NCAH from PP. Conclusion Basal 17OHP provides excellent diagnostic accuracy for 21OHD-CAH in children with PP. A dual-threshold strategy is supported: 170 ng/dL (5.1 nmol/L) as optimal screening cutoff and 410 ng/d (12.4 nmol/L) as highly specific diagnostic threshold, obviating ACTH testing. These genotype-anchored, RIA-derived cutoffs are readily applicable in clinical settings using comparable assays.