The purpose of this study was to determine possible cut-off levels of basal DHEA-S percentile rank in the differential diagnosis of patients with Cushing's syndrome (CS) with ACTH levels in the gray zone and normal DHEA-S levels. In this retrospective study including 623 pathologically confirmed CS, the DHEA-S percentile rank was calculated in 389 patients with DHEA-S levels within reference interval. The patients were classified as group 1 (n=265 Cushing's disease; CD), group 2 (n=104 adrenal CS) and group 3 (n=20 ectopic ACTH syndrome).ROC-curve analyses were used to calculate the optimal cut-off level of DHEA-S percentile rank in the reference interval in the differential diagnosis of CS, and the effectiveness of this cut-off level in the identification of the accurate etiology of CS was assessed in patients who were in gray zone according to their ACTH levels. The DHEA-S percentile rank in the reference interval were significantly lower in group 2 compared to the other two groups (p<0.001), while group 1 and group 3 had similar levels. The optimal cut-off level of DHEA-S percentile rank in the reference interval providing differential diagnosis between group 1 and group 2 was calculated as 19.5th percentile (80.8% sensitivity, 81.5% specificity) and the level demonstrated the accurate etiology in 100% of CD and 76% of adrenal CS patients who were in the gray zone. This study showed that the cut-off value of DHEA-S level less than 20% of the reference interval could be used for differential diagnosis of CD and adrenal CS with high sensitivity and specificity, and it should be taken into the initial evaluation.
Thieme. All rights reserved.