Vasopressin responders have a high prevalence of hypertension and insulin resistance in spite of a lower secretion of cortisol in subclinical Cushing's syndrome

The hyper-responsiveness of plasma cortisol to vasopressin has been reported in patients with ACTH-independent bilateral macronodular adrenocortical hyperplasia (AIMAH) and in certain cases of Cushing's adenomas with overexpression of V1a receptors in the adrenal gland. We previously reported a high prevalence of the adrenal hyper-responsiveness of plasma cortisol to vasopressin (vasopressin responders) in Japanese patients with Cushing's syndrome (CS), in which vasopressin responders had low autonomous cortisol secretion and were diagnosed as having subclinical CS. In the present study, we aimed to evaluate the pathophysiological significance of vasopressin responsiveness in patients with subclinical CS in terms of hormonal secretion and metabolic abnormalities. We compared 14 vasopressin responders and 13 non-responders with subclinical CS, admitted to Chiba University Hospital between 1999 and 2007. Among these patients, the vasopressin responders were found to have lower plasma cortisol levels than non-responders at midnight (P<0.05) and in the morning following overnight dexamethasone suppression at 1 mg (P<0.05). On the other hand, they showed a significantly higher frequency of hypertensive complications than non-responders. In addition, according to 75g-OGTT the vasopressin responders had significantly higher levels of plasma insulin at 60 min than the non-responders, and their cortisol response to vasopressin had a significantly positive association with HOMA-IR (r=0.671, P<0.05). A high expression level of V1a receptor mRNAs was observed in the resected adrenal glands of vasopressin responders with AIMAH and unilateral adenomas. Though vasopressin responders with subclinical CS had lower autonomous cortisol secretion, they had a high prevalence of hypertension, in which insulin resistance was closely correlated with cortisol response to vasopressin.