Hypothalamic-pituitary adrenal function in end-stage non-alcoholic liver disease

Abstract Patients with end-stage liver disease have significant mortality often associated with intercurrent episodes of bleeding or sepsis. Intact adrenal function is essential in such situations. In order to test the hypothesis that adrenal insufficiency might be present in severe liver disease, hypothalamic-pituitary adrenal function was evaluated in patients with end-stage liver disease awaiting transplantation. The study had a prospective, open comparative design with patients restricted to those having non-alcoholic liver disease in order to avoid the confounding direct effects of alcohol on adrenocortical function. Fifty-one consecutive patients with end-stage, non-alcoholic liver disease undergoing evaluation for liver transplantation and 40 healthy controls were studied. Patients who had used corticosteroids (n= 8) or who were unable to complete the investigations (n= 5) were excluded leaving 38 patients eligible for analysis. Adrenal function was evaluated under basal conditions by single morning measurements of plasma total and free cortisol, corticosteroid-binding globulin, dehydroepiandrosterone sulfate and by adrenal stimulation indirectly using insulin-induced (0.1 U/kg, i.v.) hypoglycaemia and/or directly by adrenocorticotrophic hormone (ACTH); 250 μg tetracosactrin, i.v.) stimulation. Compared with healthy controls, patients with liver disease had a 64% reduction in maximal increments of plasma cortisol to indirect adrenal stimulation via insulin-induced hypoglycaemia and a 39% reduction to direct adrenal stimulation by ACTH (all P < 0.001). There was a significant negative correlation between the severity of underlying liver disease as assessed by Child-Pugh scores and peak control responses to ACTH (r= -0.647, P < 0.0001) and insulin-induced hypoglycaemia (r= -0.597, P < 0.0001). Patients with liver disease also exhibited significantly blunted and delayed hypoglycaemic responses to insulin (0.1 U/kg), brisker growth hormone responses of reduced magnitude and decreased corticosteroid-binding globulin levels. Baseline morning cortisol, free cortisol and dehydroepiandrosterone sulfate levels were unchanged compared with healthy controls. Patients on spironolactone had lower basal and peak cortisol responses to ACTH and hypoglycaemia, but the reductions were unrelated to spironolactone dose. Although insulin resistance and spironolactone therapy are confounding factors, it can be concluded that hypothalamic-pituitary regulation of adrenal function is defective in end-stage non-alcoholic liver disease. It is therefore possible that functional central adrenal insufficiency might contribute to the mortality of patients with end-stage liver disease and raises the question of the need for controlled studies of adrenocortical replacement therapy during acute deteriorations (sepsis and haemorrhage) in severe hepatic disease.