Characterising adrenal function using directly measured plasma free cortisol in stable severe liver disease

Abstract

Adrenal insufficiency (AI) has been reported in patients with advanced liver disease. Diagnosing AI is problematic owing to controversies in using total serum cortisol as a measure of adrenal function. No published data exist on directly measured plasma free cortisol (PFC) in patients with liver disease. This prospective study compared serum total and measured plasma free cortisol to evaluate adrenal function in clinically stable cirrhotic patients and healthy controls. Cortisol levels were measured at baseline and following 250 μg corticotrophin. AI was defined by total cortisol increments (delta cortisol) of less than 250 nmol/L, or a peak total cortisol under 500 nmol/L after cosyntropin. We used a peak plasma free cortisol concentration of 33 nmol/L as the threshold for AI. Forty-three consecutive patients and 10 healthy controls were studied. Cirrhotic patients had significantly lower peak (526 vs. 649 nmol/L, p=0.004) and delta total cortisol (264 vs. 397 nmol/L, p = 0.002) responses compared to healthy controls. However, basal plasma free cortisol was higher in patients (10.9 vs. 6.4 nmol/L, p = 0.03), and there were no differences in peak plasma free cortisol (p = 0.69) between the two groups. The prevalence of AI using total cortisol criteria was 58% compared to 12% using free cortisol (p<0.001). In patients with stable severe liver disease, a significant discrepancy exists between the rates of diagnosis of AI using the total and free cortisol criteria. We would advise caution in the interpretation of adrenal function testing using total cortisol measurements in this group.