The pituitary-adrenal axis is activated more in non-survivors than in survivors of cardiac arrest, irrespective of therapeutic hypothermia

OBJECTIVE: To investigate the effect of therapeutic hypothermia in the prognostic value of the pituitary-adrenal axis in comatose patients after cardiac arrest.

DESIGN: Prospective observational study in intensive care units (ICU) of a university and an affiliated regional hospital.

PATIENTS: Twenty-nine consecutive patients, in coma after cardiac arrest, admitted to the ICU and treated by hypothermia.

MEASUREMENTS: On ICU-admission (T=1), at reaching the target of 32-33 degrees C during therapeutic hypothermia (T=2), at the end of hypothermia (T=3) and 48h later (T=4), plasma adrenocorticotrophic hormone (ACTH), serum cortisol, albumin and corticosteroid-binding globulin (CBG) were measured. A short 250 microg ACTH test was performed at each time-point, except at T=1. The free cortisol index (FCI) and free cortisol calculated by Coolens method were also evaluated.

RESULTS: The ICU mortality was 59%, including withdrawal of life-sustaining treatment in 45% because of negative somatosensory evoked potentials. ACTH and (free) cortisol levels (mean 13.1 pmol/L vs. 6.0 pmol/L and 1250 nmol/L vs. 596 nmol/L, respectively) were higher in non-survivors than in survivors. Levels decreased in time, but the relative difference between outcome groups was maintained until T=4. The cortisol response to ACTH was lower in non-survivors at T=3 (P=0.047) only.

CONCLUSIONS: In comatose patients resuscitated from cardiac arrest, the pituitary-adrenal axis is activated particularly in those dying in the ICU, irrespective of therapeutic hypothermia. Hence, activation of the axis may be a marker of fatal cerebral damage. There is no firm evidence for relative adrenal insufficiency associated with death and a transiently blunted cortisol response to ACTH in non-survivors may be attributed to higher baseline values.