Utility of plasma dehydroepiandrosterone sulphate determination in adrenal incidentalomas

To evaluate whether low DHEA-S levels are predictors of cortical origin, benignity and hormonal activity in incidentally detected adrenal masses, thirty-five patients with adrenal incidentalomas were studied. All patients were operated on and the diagnosis was histologically confirmed. Basal endocrine workup included plasma determination of cortisol before and after dexamethasone (1 mg overnight), plasma ACTH (08:00 h), 17-OH-progesterone, testosterone and potassium, standing plasma renin activity and aldosterone, supine and standing plasma noradrenaline and adrenaline. If necessary, we performed dexamethasone suppression tests at low (2 mg) and high (8 mg) doses, or the loperamide test (16 mg os) for evaluation of glucocorticoid activity and the glucagon test (1 mg i.v.) for exploring adrenal medulla function. Plasma DHEA-S was measured in all patients and the results were compared to those obtained in controls matched for age, sex and menopausal status. Suppression of DHEA-S was found in 11 out of 35 patients (31.5%). However, this hormonal finding occurred in 50% of the extracortical adrenal lesions, while in proven cortical adenomas (no. = 19) it was detected in only 5 patients (26.3%). Sensitivity, specificity, diagnostic accuracy and positive predictive value of low DHEA-S in indicating a cortical origin of the mass were 0.27, 0.0, 0.25, and 0.80. In malignancies (no. = 6) low DHEA-S levels were found in 1 out of 2 metastases and never in cortical carcinomas. Sensitivity, specificity, diagnostic accuracy and positive predictive value of low DHEA-S in indicating a benign form were 0.34, 0.83, 0.42, and 0.91. Six out of 19 patients with cortical adenomas showed signs of hypothalamic-pituitary adrenal (HPA)-axis dysfunction. Low DHEA-S levels were found in 50% of adenomas with HPA-axis abnormality and in 15.3% of adenomas without hormonal activity. Sensitivity, specificity, diagnostic accuracy, and positive predictive value of low DHEA-S levels in indicating hormonal activity of the mass were 0.50, 0.84, 0.73, and 0.60. Our data indicate that the association between low DHEA-S levels and adrenal incidentalomas is frequent. Low DHEA-S appears to be a poor predictor of hormonal activity with low sensitivity and specificity in respect of cortical origin and benignity of the mass. In conclusion, our results show that DHEA-S measurement does not offer relevant clinical information in the management of adrenal incidentalomas.