Hepatoadrenal Syndrome – Can Serum Free Cortisol Concentration Free Us from Empiric Glucocorticoids?

Adrenal insufficiency (AI) in the setting of cirrhosis is often considered, but is insufficiently characterized. Abnormal cortisol metabolism has been attributed to  impaired processing of cholesterol precursors, altered 11β-HSD activity, and inflammatory cytokines. Diagnosis of clinical AI remains uncertain, and often relies on failure to increase serum cortisol above assay-specific thresholds after cortrosyn stimulation. Given the high prevalence of hypoalbuminemia in cirrhosis, low cortisol binding globulin (CBG) is expected, likely invalidating published thresholds for serum cortisol.

Case Description :

We describe 9 consecutive hospitalized men and women with cirrhosis aged 37-63. Endocrinology was consulted for evaluation of potential AI. MELD scores ranged from 7-34, and three subsequently underwent liver transplantation. Cortisol testing was typically performed because of hypotension and hyponatremia. The patients had morning cortisol values between 2.1-9.0, with mean 5.1±2.3 (reference 6.0-18.4 ug/dL). Post-cortrosyn stimulation cortisol concentrations ranged from 8.7-12.3 mg/dL, leading to initiation of empiric glucocorticoid replacement by medicine teams. CBG was low in both patients in whom it was assessed. Nadir albumin levels ranged from 1.5-2.9 g/dL. Free serum cortisol levels ranged from 0.26-1.79 ug/dL (reference 8AM: 0.21-1.04 ug/dL) by equilibrium dialysis LC/MS/MS analysis.

Discussion :

We present data from 9 patients with cirrhosis, suspected of having AI, in whom free serum cortisol remained in reference range despite low total serum cortisol levels. AI in cirrhosis, often referred to as hepatoadrenal syndrome, has previously been described in up to 30% of cirrhotic patients. Hypotension and hyponatremia are clinical features common to both AI and cirrhosis. Cortisol secretion and free cortisol clearance have been shown to be reduced in advanced cirrhosis, but correlation with clinical outcomes is unclear. Additionally, low affinity CBG mutations and estrogen therapy (which increases CBG) alter total cortisol while free cortisol remains in reference range. CBG and free cortisol labs typically take longer to result than total cortisol levels, and treatment with empiric glucocorticoids is often employed. This practice risks infection, hyperglycemia, and hypervolemia. Our case series suggests that glucocorticoids should be used judiciously in the cirrhotic population, as many patients with derangements in CBG and total cortisol still exhibit free cortisol values within reference range. In this context, free cortisol levels may provide a better guide for glucocorticoid replacement and/or withdraw than metrics relying on total cortisol.