The Hypothalamus-pituitary-adrenocortical Response to Critical Illness: A Concept in Need of Revision

Abstract

Based on insights obtained during the past decade, the classical concept of an activated hypothalamus-pituitary-adrenocortical axis in response to critical illness is in need of revision. After a brief central hypothalamus-pituitary-adrenocortical axis activation, the vital maintenance of increased systemic cortisol availability and action in response to critical illness is predominantly driven by peripheral adaptations rather than by an ongoing centrally activated several-fold increased production and secretion of cortisol. Besides the known reduction of cortisol-binding proteins that increases free cortisol, these peripheral responses comprise suppressed cortisol metabolism in liver and kidney, prolonging cortisol half-life, and local alterations in expression of 11βHSD1, glucocorticoid receptor-α (GRα), and FK506 binding protein 5 (FKBP51) that appear to titrate increased GRα action in vital organs and tissues while reducing GRα action in neutrophils, possibly preventing immune-suppressive off-target effects of increased systemic cortisol availability. Peripherally increased cortisol exerts negative feed-back inhibition at the pituitary level impairing processing of pro-opiomelanocortin into ACTH, thereby reducing ACTH-driven cortisol secretion, whereas ongoing central activation results in increased circulating pro-opiomelanocortin. These alterations seem adaptive and beneficial for the host in the short term. However, as a consequence, patients with prolonged critical illness who require intensive care for weeks or longer may develop a form of central adrenal insufficiency. The new findings supersede earlier concepts such as "relative," as opposed to "absolute," adrenal insufficiency and generalized systemic glucocorticoid resistance in the critically ill. The findings also question the scientific basis for broad implementation of stress dose hydrocortisone treatment of patients suffering from acute septic shock solely based on assumption of cortisol insufficiency.

Keywords: binding proteins; cortisol; critical illness; glucocorticoid receptor; sepsis.

Graphical Abstract

Figure 1.

The relative change over time in plasma hormone concentrations of the HPA axis and the central adaptations explaining the prolonged critically ill phenotype. After a brief rise, plasma ACTH becomes suppressed and returns to low-normal levels in patients with prolonged critical illness, whereas plasma concentrations of POMC are elevated throughout the stay in the ICU. Plasma total and free cortisol concentrations are elevated on admission to the ICU and decrease progressively with time in the ICU, which results in the absence of hypercortisolemia when patients are in the ICU for weeks or longer. Sustained central activation of the corticotrope cells increases CRH- and AVP-mediated POMC expression. The sustained peripherally driven high systemic cortisol levels, however, through increased corticotrope GRα ligand-binding, suppress expression of PC1/3, the enzyme controlling POMC to ACTH processing, and increase expression of Annexin A1, the negative regulator of mature ACTH secretion, together causing low plasma ACTH. As a result, corticotrope POMC may accumulate and subsequently leak into the circulation through the constitutive pathway, explaining the high plasma POMC concentrations. In prolonged critical illness, the suppressed ACTH processing and impaired ACTH secretion may explain the suppressed incremental ACTH response to a CRH stimulation test. The decline in total and free cortisol in these patients with prolonged critical illness treated in the ICU may thus reflect a central HPA axis suppression representing a risk of adrenal insufficiency. In survivors, retested 1 week after ICU discharge, a rebound ACTH and cortisol rise is present which suggested reversibility of the central HPA axis suppression. Created with Biorender.com.

Figure 2.

Schematic overview of the main, mostly peripheral, drivers of the increased systemic cortisol availability in response to critical illness. Plasma free cortisol concentrations are high because of reduced cortisol plasma binding proteins and suppressed cortisol breakdown in liver and kidney, allowing increased distribution of cortisol to central and peripheral target tissues. The daily amount of produced cortisol is only moderately increased or normal to which increased plasma POMC could contribute in the absence of elevated plasma ACTH. Tissue-specific adaptations in GRα signaling safeguard increased cortisol action in vital organs and tissues, whereas in neutrophils, cortisol action is suppressed whereby off-target immune suppressive effects of increased systemic cortisol availability could be at least partially prevented. Created with Biorender.com.