The hypothalamic-pituitary-adrenal axis of patients with severe sepsis: Altered response to corticotropin-releasing hormone

OBJECTIVETo investigate the functional integrity of the hypothalamic-pituitary-adrenal (HPA) axis in patients with severe sepsis by stimulating with corticotropin-releasing hormone (CRH). DESIGNProspective observational study in consecutive intensive care unit patients with severe sepsis. SETTINGSurgical intensive care unit and outpatient department of endocrinology in a university hospital. PATIENTSThe study included 20 patients with the diagnosis of severe sepsis; six critically ill, nonseptic patients after major surgery; ten patients with primary adrenal insufficiency; ten patients with anterior pituitary insufficiency; and ten individuals without clinical signs of HPA axis disturbance. INTERVENTIONSCRH tests were performed with an intravenous bolus injection of 100 μg of human CRH. MEASUREMENTS AND MAIN RESULTS We studied the functional integrity of the HPA axis in patients with severe sepsis by performing the CRH test. In addition, during the period of severe sepsis, we repeatedly measured basal plasma concentrations of adrenocorticotropin hormone (ACTH) and cortisol. The mean basal plasma cortisol concentration was decreased significantly in nonsurvivors with severe sepsis (288.8 ± 29.1 [sem] nmol/L) compared with survivors (468.1± 18.6 nmol/L;p < .01). By calculating the ACTH/cortisol indices, we found no evidence for adrenal insufficiency in patients with severe sepsis. The mean ACTH/cortisol indices of nonsurvivors with severe sepsis (0.02 ± 0.008) and survivors (0.01 ± 0.002) were significantly lower compared with the index of patients with primary adrenal insufficiency (6.8 ± 1.0;p < .001). In contrast, in nonsurvivors with severe sepsis, the plasma cortisol response to CRH stimulation was impaired compared with survivorsThe mean basal cortisol concentration within the CRH test was 269.4 ± 39.8 nmol/L in nonsurvivors compared with 470.8 ± 48.4 nmol/L in survivors and increased to a peak value of 421.6 ± 72.6 nmol/L in nonsurvivors and 680.7 ± 43.8 nmol/L in survivors (p < .02). However, the change in plasma cortisol, expressed as mean ± sem and calculated by subtracting the basal cortisol from the peak cortisol after CRH stimulation, was not significantly different in survivors with severe sepsis (243.5 ± 36.1, range 111.0–524.0 nmol/L, n = 15) compared with nonsurvivors (161.0 ± 38.9, range 42.0–245.0 nmol/L, n = 5;p > .05). CONCLUSIONSWe found lower basal plasma cortisol concentrations in nonsurvivors compared with survivors of severe sepsis.