These findings confirm that changes of plasma leptin concentration occur in PD patients with loss of weight.

Fifty patients bilaterally adrenalectomized for Cushing's disease were followed for 1 to 22 (mean, 12) years. In 14 of them (28%) Nelson's syndrome appeared within 1.5 to 12 (mean, 4.8) years after adrenalectomy. All the patients were deeply pigmented, 12 had a radiologically demonstrable tumour and six had visual defects. Plasma ACTH levels ranged from 450 to 8000 ng/l. However in every case at least one estimation during circadian studies equalled 2000 ng/l. One patient with an anaplastic pituitary tumour died 3 years after the discovery of the tumour. Anaplasia was also diagnosed in another patient with recurrence of pituitary tumour after a transsphenoidal operation. Symptoms of tumour infarction occurred in two patients, followed by clinical remission of Nelson's syndrome. In the majority of cases the course was benign. We conclude that all cases, however, should be followed indefinitely because Nelson's tumours are not infrequently aggressive.

MRI, ophthalmologic examination and plasma ACTH determination were the most valuable investigations for early diagnosis of Nelson's syndrome. Early neurosurgery offered the best outcome in our group of patients.

Excessive GH response to various stimuli has been frequently described in diabetes mellitus. We studied the GH response to a synthetic GHRH in a group of 16 non-obese Type I diabetic patients. GHRH (1\p=n-\44) given as iv bolus at a dose of 50 \ g=m\ g induced a markedly greater GH response in the diabetic than in the normal subjects with peak values of 39.5 and 14.7 \g=m\g/l,respectively, and the differences were significant from 15 to 60 min. Peak GH level was 44.6 \g=m\g/l in diabetic patients without retinopathy and 34.2 \g=m\g/l in patients with retinopathy, but the difference was not significant. Peak GH levels did not correlate with metabolic control of disease estimated by basal glucoseand HbA1 levels nor with age, weight of the patients, and duration of the disease. It is concluded that Type I diabetic patients show an exaggerated GH response to GHRH and this response does not correlate with the metabolic control of diabetes.

Abstract. The role of a high CRH level in normal pregnancy remains unknown. Therefore we evaluated the concentrations of CRH and the related hormones in patients with pregnancy-induced hypertension. Fourteen women with pregnancy-induced hypertension, aged 20-39, at 30-39 gestational week, were investigated. The control group consisted of 20 healthy pregnant women matched according to gestational age. Plasma CRH, β-endorphin-like immunoreactivity, cortisol, and human placental lactogen were measured by radioimmunoassay, ACTH by an immunoradiometric method. It was found that in hypertensive patients the mean CRH concentration was significantly higher (4257±840 (sem) ng/l) than that in healthy pregnant women (1083±227 ng/l, p<0.001). The concentration of ACTH, however, was only slightly higher 65.0±6.0 vs 50.7±2.5 ng/l, p<0.025, whereas the differences in β-endorphin, cortisol and human placental lactogen were not significant. In both groups there was no correlation between the CRH level and those of the related hormones. In healthy pregnant women the CRH level closely correlated with gestational age (r=0.76, p<0.001), whereas in patients with hypertension no such correlation was present (r=0.29). We assume that the marked enhancement of plasma CRH in pregnancy-induced hypertension is probably caused by its decreased breakdown in ischemic placental tissue, but its increased synthesis in the placenta and its indirect counterregulatory hypotensive role must also be considered.