Endocrine Abstracts

Critical illness is hallmarked by hypercortisolemia, traditionally attributed to stress-induced hypothalamic–pituitary–adrenal axis activation. However, as low plasma ACTH concentrations and reduced responsiveness to ACTH have been reported, we hypothesized that reduced cortisol metabolism could play a major role. To systematically test this hypothesis, we recently performed six studies in matched ICU patients and healthy controls. These measured daily ACTH and cortisol plasma concentrations (n=59); morning plasma cortisol clearance, metabolism and production via deuterated-steroid tracer infusions (n=20); plasma clearance of 100 mg hydrocortisone bolus injection (n=28); urinary cortisol metabolites (n=51); mRNA and protein of major cortisol-metabolizing enzymes in liver and adipose tissue (n=64); and repeated (every 10 min) plasma ACTH and cortisol concentrations overnight to quantify pulsatile hormone secretion and dose-responses (n=48). Morning total and free plasma cortisol concentrations were consistently higher, while ACTH was lower in critically ill patients than in healthy controls. Morning cortisol production was only 83% higher than in healthy controls, and correlated positively with pro-inflammatory cytokines, not with ACTH. Patients without systemic inflammation did not have an elevated cortisol production at all despite very high plasma cortisol. In contrast, cortisol clearance during low dose tracer infusion and after 100 mg hydrocortisone was uniformly lowered to less than half the values in healthy subjects, irrespective of the inflammatory status, substantially contributing to the 3.5-fold increased plasma cortisol. Impaired cortisol clearance also correlated with lower cortisol response to ACTH injection. Reduced cortisol metabolism was due to reduced inactivation of cortisol by 5β- and 5α-reductases in liver and by 11β-HSD2 in kidney, as suggested by urinary steroid ratios, tracer kinetics and enzyme expression in liver biopsies. Reduced expression of cortisol metabolizing enzymes in liver biopsies was associated with higher plasma bile-acids and cortisol. Hypercortisolism during critical illness coincided with suppressed nocturnal pulsatile ACTH and cortisol secretion and with a normal ACTH/cortisol dose–response. In conclusion, during critical illness, reduced cortisol breakdown contributes to hypercortisolemia, and ACTH suppression. This may be important not only to increase circulating cortisol levels but also to potentiate cortisol concentrations and activity within those vital tissues that express the metabolizing enzymes. The data also suggest that ‘stress-doses’ of hydrocortisone (200 mg/day), advocated to replace cortisol production in critically ill patients presumed to suffer from adrenal failure, may be at three- to sixfold too high.