Endocrine Abstracts

The hypothalamo-pituitary-adrenal (HPA) axis is essential component of the physiological stress response. Functional hypoadrenalism is a hypothetical contributor to mortality in the critically ill, though biochemical evidence for this is conflicting. Most studies have shown that 'high' random plasma cortisol in the first 24 hours of illness is associated with a high mortality. However, these data may be confounded by increased severity of illness in high mortality cohorts. Given this background, we have sought to identify differences in on-admission plasma cortisol (OPC) between patients matched for severity of illness, but differing in outcome (survival vs. non-survival) in a novel, prospective study: an essential step in testing the hypothesis of functional hypoadrenalism in the critically ill.
APACHE II predicted mortality was used to prospectively divide a cohort of ITU patients into low, medium and high tertiles of predicted mortality (0-32.99%, 33-65.99%, 66-100%). OPC was assayed (ADVIA Centaur Immunoassay, inter and intra assay variation 1.86-5.45 % and 2.96-3.69 %) within 5 hours of admission. Study protocols were approved by appropriate local ethics review.
73 (18 non-survivors) were recruited from 125 eligible. Exclusion criteria included those with conditions not appropriate for APACHE II modelling, and previous or concurrent exogenous steroid treatment. OPC was higher in non-survivors (812 vs 533 nmol/l, p=0.047, Mann-Whitney U). After correction for severity of illness, no differences in OPC (median; IQR) were apparent between groups with different outcomes in low (510; 435 vs 572; 402, p=0.557), medium (461;1010 vs 1195.5; 1270, p=0.136) or high (757;774 vs 1044;1295, p=0.289) predicted mortality tertiles. We believe this to be a novel approach to differentiating between the patho-physiological, physiological and/or epiphenomological nature of HPA changes in critical illness, and thus identify a sub-group who may benefit from treatment.