Abstract:Objective: In ICU patients, abnormal liver tests are common. Markers of cholestasis are associated with adverse outcome. Research has focused on the possibility that mild hyperbilirubinemia, instead of indicating inadvertent cholestasis, may be adaptive and beneficial. These new insights are reviewed and integrated in the state-of-the-art knowledge on hepato-biliary alterations during sepsis and other critical illnesses.Data Sources: Relevant publications were searched in Medline with search terms bile, bile acids, cholestasis, critical illness, intensive care, sepsis, alone or in combination.

Sepsis is hallmarked by hypercortisolemia, a stress response essential for survival. This elevation in plasma cortisol is partially brought about by suppressed hepatic cortisol breakdown. We demonstrate that a controlled downregulation of the hepatic glucocorticoid receptor (hepatic GR) is crucial. In a mouse model of fluid-resuscitated, antibiotic-treated abdominal sepsis and in human intensive care unit patients, sepsis reduced hepatic GR expression and signaling but increased (free) plasma cortisol/corticosterone, explained by suppressed cortisol/corticosterone-binding proteins and A-ring reductases. However, further experimental inhibition of hepatic GR with short hairpin RNA (shRNA) in septic mice increased mortality fivefold. Acutely, this further hepatic GR suppression prevented the rise in total corticosterone but further reduced binding proteins, resulting in elevated free corticosterone. After 3 days of shRNA-GR inhibition in sepsis, both total and free corticosterone levels were elevated, now explained by an additional reduction in A-ring reductase expression. Hepatic GR inhibition blunted the hyperglycemic stress response without causing hypoglycemia but also markedly increased circulating and hepatic inflammation markers and caused liver destruction, the severity of which explained increased mortality. In human sepsis, glucocorticoid treatment further suppressed hepatic GR expression, which could directly predispose to worse outcomes. In conclusion, sepsis partially suppressed hepatic GR expression, which appeared crucial to upregulate free cortisol/corticosterone availability. However, further sustained hepatic GR suppression evoked lethal excessive liver and systemic inflammation, independent of systemic cortisol/corticosterone availability.

Liver dysfunction and jaundice are traditionally viewed as late features of sepsis and other critical illnesses and are associated with a complicated ICU stay. However, study results suggest that cholestatic alterations occur early in the course of critical illnesses, perceived only as minor abnormalities in routinely used biochemical liver tests. Inflammation-induced alterations in the transport of bile acids (BAs) appear to drive BAs and bilirubin toward the systemic circulation. Ongoing BA synthesis with an, at least partial, loss of feedback inhibition further contributes to elevated circulating BAs and bilirubin. To what extent these changes reflect a biochemical epiphenomenon, true illness-induced liver dysfunction, or a beneficial and adaptive response to illness should be investigated further. Because of the lack of specificity of standard laboratory tests, especially in the context of a complex systemic condition such as critical illness, identifying true cholestatic liver dysfunction remains a great challenge. However, high levels of cholestatic markers that are sustained in patients with prolonged critical illness almost always indicate a complicated illness course and should be monitored closely. Preventing cholestatic liver dysfunction comprises minimizing inflammation and hypoxia in the liver and preventing hyperglycemia, avoiding early use of parenteral nutrition, and reducing the administration of avoidable drugs. Future research on the effects of BAs and on modulating underlying drivers of cholestasis induced by critical illness is warranted as this could open perspectives for a targeted diagnostic approach and ultimately for novel therapies to improve outcome.

Supplemental Digital Content is available in the text

Supplemental Digital Content is available in the text.