VOLUME 118, ISSUE 2, P384-391
Grethe Å. Ueland, Ph.D., Sandra R. Dahl, Ph.D., Paal Methlie, Ph.D., Saleh Hessen, M.D., Eystein S. Husebye, Professor, Per M. Thorsby, M.D., Ph.D.
To define liquid chromatography tandem mass spectrometry (LC-MS/MS)–based cutoff levels and panels of steroid hormones, to improve diagnosis of nonclassic congenital adrenal hyperplasia (NCCAH) and other partial enzyme defects in the adrenals.
Prospective cohort analysis.
University hospital-based tertiary endocrine center.
One hundred and twenty-one healthy adults and 65 patients evaluated for possible NCCAH (validation cohort).
The LC-MS/MS–determined cutoffs for 11 steroids (basal and cosyntropin-stimulated) were defined by 2.5% and 97.5% percentile in healthy subjects. Validation cohort was used for comparison.
Main Outcome Measures
Percentage of patients diagnosed with NCCAH among patients with polycystic ovary syndrome (PCOS)-like symptomatology. Evaluation of the defined LC-MS/MS–based cutoff levels for steroid hormones among this patient group.
Of the 65 PCOS-like patients evaluated for possible NCCAH, 8 (12.5%) were discovered and genetically verified, and 2 had classic congenital adrenal hyperplasia. Cosyntropin-stimulated 17-hydroxyprogesterone (17OHP) showed the best diagnostic accuracy for NCCAH with an area under the curve of 0.95 (0.89–1.0 with a sensitivity of 86% and a specificity of 88%. In homozygote patients, 21-deoxycortisol and 17OHP levels were elevated, in heterozygote patients only 17OHP (basal or stimulated) was raised. Four healthy patients in the validation cohort had 17OHP above the basal cutoff.
The NCCAH syndrome is frequent in patients with suspected PCOS, and should be considered as a routine screening when assessing infertility. We suggest the use of serum steroid profiling, including 21-deoxycortisol, together with the cosyntropin stimulation test with 17OHP. Our data support a 17OHP cutoff of 8.5 nmol/L (2.8 ng/mL) 60 minutes after cosyntropin stimulation, when measured with LC-MS/MS, significantly lower than current European guidelines.
Clinical trials Number