The hypothalamic-pituitary-adrenal axis was investigated in all six patients requiring glucocorticoid replacement 2.5\p=n-\11 years after unilateral adrenalectomy for adrenal adenomas causing Cushing's syndrome. The hypothalamic-pituitary-adrenal axis was assessed by insulin induced hypoglycaemia and CRF testing in each patient. Two patients showed normal cortisol and ACTH responses to hypoglycaemia. Two patients showed subnormal cortisol responses to hypoglycaemia in the presence of high or normal basal ACTH concentrations. ACTH concentrations increased with both hypoglycaemia and CRF. Two patients showed subnormal cortisol responses to hypoglycaemia and CRF. One of these patients showed an ACTH rise following hypoglycaemia but not CRF. Defects at either hypothalamic-pituitary or adrenal levels were demonstrated and recovery of the axis appears to commence at the hypothalamic-pituitary level.Surgery for adrenal adenomas causing Cushing's syndrome is followed by a period of hypothalamic-pituitary-adrenal ( ) suppression of vari¬ able duration (Kepler et al. 1948;Cecil 1933;Rapaport et al. 1952). Although most patients eventually achieve recovery of adrenal function, series reporting the long-term results of adrenal surgery often include a small number of patients in whom prolonged follow-up fails to demon¬ strate such recovery (Welbourn et al. 1971;Valimaki et al. 1984). The mechanisms resulting in continued suppression of adrenocortical activity in these patients are little understood. We investi¬ gated the axis in six patients requiring glucocorticoid replacement 2.5 -11 years after unilateral adrenalectomy by assessing the cortisol and ACTH responses to insulin induced hypogly¬ caemia ( ) and intravenous injection of human CRF.
Patients and MethodsSix patients requiring glucocorticoid replacement ther¬ apy 2.5 -11 years (mean 5.2 years) following unilateral adrenalectomy for Cushing's syndrome were investi¬ gated. Previous unsuccessful attempts to stimulate adreno-cortical secretion with prolonged courses of depot tetracosactrin had been made in 5 patients. De¬ tails of patients and previous treatment are shown in Table 1. Local ethical committee permission for this project had been obtained. After informed consent had been obtained, each patient was investigated on two occasions separated by at least 24 h. On both occasions, hydrocortisone replacement therapy had been withdrawn for 24 h prior to insulin (0.15 U/kg) induced hypogly¬ caemia (glucose < 2.2 nmol/1). Blood samples were withdrawn at 0, 30, 45, 60, 90, 120 min for cortisol, glucose and ACTH assays.On the second occasion, human CRF (100 µg) in acid saline pH 3.0, stored at -70°C was reconstituted in normal saline and injected iv to recumbant patients over 60 sec. Blood was withdrawn at 0, 15, 30, 45, 60, 90, 120 min for cortisol and ACTH assays.Normal data was derived from healthy adult volun¬ teers (10 for insulin induced hypoglycaemia (IIH), 6 for CRF testing). Serum glucose was estimated by the glucose oxidase method.