ESPE Abstracts (2014) 82 P-D-1-2-1

Allelic Frequencies of CYP21A2 Variants and Genotype-Phenotype Correlations in a Cohort of 660 CAH Patients from Germany and Austria

Stefan Riedla,b, Friedrich-Wilhelm Röhlc, Susann Emptingd, Walter Bonfige, Helmuth-Günther Dörrf, Reinhard Hollg & Klaus Mohniked


aSt Anna Children’s Hospital, Medical University of Vienna, Vienna, Austria; bDepartment of Pediatric Pulmology, Allergology, and Endocrinology, Medical University of Vienna, Vienna, Austria; cInstitute of Biometry and Medical Informatics, Otto-von-Guericke University Magdeburg, Magdeburg, Germany; dDepartment of Pediatrics, Otto-von-Guericke University Magdeburg, Magdeburg, Germany; eDepartment of Pediatrics, Technical University Munich, Munich, Germany; fDepartment of Pediatrics and Adolescent Medicine, Friedrich-Alexander University Erlangen-Nuremberg, Erlangen, Germany; gInstitute of Epidemiology and Medical Biometry, University of Ulm, Ulm, Germany


Background: Congenital adrenal hyperplasia (CAH) due to a CYP21A2 defect (autosomal recessive) leads to salt wasting (SW), simple virilizing (SV), or non-classical (NC) phenotypes basically depending on residual 21-hydroxylase (21-OH) function on the least affected allele.

Objective and Hypotheses: To test prediction of CAH phenotype based on genotype classification.

Method: Patient data from 37 centers were retrieved from a central data base as part of a German quality assurance program (AQUAPE) within the German Association for Pediatric Endocrinology and Diabetes (DGKED). Allelic frequency and distribution of the 11 most common CYP21A2 mutations and deletions/conversions were analysed in 660 homozygous or compound heterozygous CAH patients. Clinical phenotypes as classified by the treating physician were compared with predicted phenotypes derived from genotype classification according to magnitude of residual 21-OH function (group null=0%; group A=0–2%; group B=2–5%; and group C=20–60%).

Results: Allelic frequency of mutations was comparable to previous studies, with deletions/conversions (29.6%) and I2G (29.2%) being the most common, followed by I172N (13.1%). Complete genotype–phenotype data sets were available in 547 patients. Severe genotypes (null and A) correlated well with clinical phenotype (SW in 97 and 91% respectively), whereas weaker genotypes (B and C) showed poor correlation with expected phenotype (SV in 45% and NC in 57% respectively), underestimating clinical severity, specifically associated with I172N and P30L. In C genotypes, this was underlined by the degree of virilization (Prader stage >1 in 28%). In A and B genotypes, a null-mutation on the second allele increased the risk of a more severe phenotype significantly.

Conclusion: In our large cohort comprising 660 patients, phenotype prediction was unreliable in weaker genotypes. Rather than clearly classifiable, CAH severity has to be regarded as a continuum requiring flexibility in clinical management.

Article tools

My recent searches

No recent searches.