Adequate adrenocortical function is essential to survive critical illness. The number of circulating eosinophils has been proposed as a marker of adrenocortical function. The goal of the present study was to determine whether eosinophilia could serve as a useful and early marker of adrenal insufficiency in critically ill patients with severe septic shock.

During a 1-year period, we studied prospectively all 294 patients admitted to our ICU. Sixteen patients (13 male/three female, 5.4% of admissions) with eosinophilia defined as more than 3% of the white blood cell count and severe septic shock, refractory to fluid and vasopressor resuscitation, were included. A high-dose (250 μg, intravenously) corticotropin stimulation test was performed in all included patients.

The mean age was 47.2 ± 18.7 years, the Acute Physiology and Chronic Health Evaluation II score on admission day was 18.6 ± 6.8 and the Sepsis-related Organ Failure Assessment score was 10.3 ± 2.7 on eosinophilia day. The mean eosinophil count was 6.9 ± 3.5% of white blood cells. Eosinophilia was present 1.9 ± 0.9 days (range 8 hours to 4 days) before the onset of septic shock. Multidrug-resistant Gram-negative bacteria in 14 patients, Gram-positive in three patients and fungi in two patients were isolated and considered responsible for sepsis. Baseline cortisol levels were 19.4 ± 8.1 μg/dl and the adrenal response to the corticotropin stimulation test was 8.3 ± 4.9 μg/dl above baseline. Eleven out of 16 patients failed to respond to the corticotropin stimulation test above the critical level of a 9 μg/dl rise, and two out of 16 patients had baseline cortisol concentration <10 μg/dl. A hydrocortisone infusion (300 mg/day) treatment resulted in haemodynamic improvement in 12 out of 16 patients (75%). The 28-day mortality (following the onset of septic shock) was 43.7%. The only independent predictor of death was age (P = 0.027).

Relative eosinophilia may be considered a useful and early bioassay for adrenocortical function assessment in critically ill patients with severe septic shock and assumed adrenocortical depression.