Non-classical 21-hydroxylase deficiency in children:: association of adrenocorticotropic hormone-stimulated 17-hydroxyprogesterone with the risk of compound heterozygosity with severe mutations

Aim: To investigate the association between levels of 17- hydroxyprogesteron e (17- OHP) and the risk of being compound heterozygous for severe mutations in children with non- classical 21-hydroxylase deficiency (NC21OHD). Methods: In 86 Spanish NC21OHD children (75 families) an analysis of the 21- hydroxylas e (21- OH) gene was performed by CYP21B- specific polymerase chain reaction amplification, allele-specific oligonucleotide hybridization and Southern blotting. Familial analysis established how the alleles segregated, and allowed the selection of 21-OH-genotyped normal and carrier children, which proved useful in determining a more precise definition of the cut- off for diagnosis. Receiver operating characteristics (ROC) curve analyses were performed to determine the potential value of 17- OHP in predicting compound heterozygosity for severe mutations. Results: Thirty- four of the 86 children (39%) were found to carry one severe 21- OH mutation (7.3% deletions or conversions, 2.7% 655G, 2.7% Q318X, 1. 3% I172N, 1.3% R356W, and 3.3% double microconversions or small conversions involving single exons). The predominant mutation was V281L (56.7%). P453S and P30L were less frequent (3. 3 and 2%). No patient showed two severe mutations. The degree of enzymic deficiency, as measured by basal or adrenocorticotropic hormone (ACTH)- stimulated 17- OHP levels in fully genotyped patients, but not clinical severity (age and number of symptoms at diagnosis), was found to be significantly greater in children with the severe/ mild genotype. ROC curve analyses revealed a strong association between ACTH- 17- OHP and genotype (area under the curve 0.908, SE 0.057). Conclusion: ACTH- stimulated 17- OHP may predict the risk of severe mutations in compound heterozygosity in children (maximum predictive value 93% sensitivity and 83% specificity for a cut- off at 151 nmol l(1)), although a certain overlap in individual values is observed and performance of molecular analysis should never be obviated in the genetic counselling of these patients.