Hypothalamic-pituitary-adrenal axis dysfunction in critically ill patients with traumatic brain injury: incidence, pathophysiology, and relationship to vasopressor dependence and peripheral interleukin-6 levels

Abstract

Objective: To investigate hypothalamic-pituitary-adrenal axis function in patients requiring mechanical ventilation for traumatic brain injury and to assess the relation of hypothalamic-pituitary-adrenal axis abnormalities with vasopressor dependence and peripheral cytokine levels.

Design: Prospective study.

Setting: General intensive care unit in a university teaching hospital.

Patients: Forty patients (33 men and 7 women) with moderate to severe traumatic brain injury (mean age, 37 +/- 16 yrs) were studied the day after termination of mechanical ventilation (7-60 days after trauma).

Interventions: First, a morning blood sample was obtained to measure baseline cortisol, corticotropin, interleukin-6, and tumor necrosis factor alpha. Subsequently, 1 microg of synthetic corticotropin was injected intravenously, and 30 mins later, a second blood sample was drawn to determine stimulated plasma cortisol. Based on data derived from healthy volunteers, patients having stimulated cortisol levels <18 microg/dL were defined as nonresponders to the low-dose stimulation test. Thirty-one patients underwent also a human corticotropin releasing hormone test.

Measurements and main results: In traumatic brain injury patients, mean baseline and low-dose stimulation test-stimulated cortisol levels were 17.2 +/- 5.4 microg/dL and 24.0 +/- 6.6 microg/dL, respectively. The median increment in cortisol was 5.9 microg/dL. Basal corticotropin levels ranged from 3.9 to 118.5 pg/mL. Six of the 40 patients (15%) failed the low-dose stimulation test. The human corticotropin releasing hormone test (performed in 26 responders and five nonresponders) revealed diminished cortisol release only in the low-dose stimulation test nonresponder patients. Corticotropin responses to corticotropin releasing hormone were consistent with both primary (three patients) and/or secondary (two patients) adrenal dysfunction. In retrospect, nonresponders to the low-dose stimulation test more frequently required vasopressors (6/6 [100%] vs. 16/34 [47%]; p =.02) and for a longer time interval (median, 0 vs. 293 hrs; p =.006) compared with responders. Furthermore, nonresponders had higher interleukin-6 levels compared with responders (56.03 vs. 28.04 pg/mL; p =.01), whereas tumor necrosis factor alpha concentrations were similar in the two groups (2.42 vs. 1.55 pg/mL; p =.53).

Conclusions: Adrenal cortisol secretion after dynamic stimulation is deficient in a subset of critically ill patients with moderate to severe head injury. This disorder is associated with prior vasopressor dependency and higher interleukin-6 levels.