Adrenal Insufficiency in the Intensive Care Unit

Relative adrenal insufficiency: This is the most common and perplexing type of adrenal insufficiency seen in patients in the ICU. Patients with relative adrenal insufficiency may present with vasopressor dependency, acute multiple organ dysfunction, hypothermia, or an inability to wean from mechanical ventilation. These patients can be identified by their limited response to adrenal stimulation tests or lower-than-expected basal cortisol levels despite critical illness.

Acute adrenal crisis or insufficiency: The acute clinical presentation typically includes profound hypotension, fever, and hypovolemia. These patients will have very low cortisol levels (<3 mcg/dL).

Chronic adrenal insufficiency

Primary adrenal insufficiency (Addison disease): The most common causes are autoimmune diseases (70%) and tuberculosis (10%). Rare causes include adrenal hemorrhage, adrenal metastasis, cytomegalovirus, human immunodeficiency virus (HIV) disease, amyloidosis, and sarcoidosis.

Secondary adrenal insufficiency: This condition is caused by inadequate production of adrenocorticotropic hormone (ACTH) due to long-term use of exogenous steroids (most common cause), hypopituitary state, or isolated ACTH deficiency.

Hyponatremia is most common.

Low levels of chloride and bicarbonate and high levels of potassium occur frequently.

Also seen are moderate eosinophilia, lymphocytosis, hypercalcemia, and hypoglycemia.

In ICU patients: The use of provocative adrenal stimulation tests in critically ill patients remains controversial. Perhaps the most widely used protocol (from Annane et al.) identifies patients with septic shock as having relative adrenal insufficiency if their baseline cortisol level is <35 mcg/dL and they respond to an ACTH stimulation test (250 mcg corticotropin) with a bump in cortisol of <10 mcg/dL (see Fig. 52-1). The patients identified as nonresponders appeared to have a reduction in mortality when given stress-dose steroids. However, in a large subsequent study, no mortality benefit with stress-dose steroids was observed for patients with or without evidence of relative adrenal insufficiency.

In non-ICU patients: In a nonstressed patient, a random cortisol level >20 mcg/dL may rule out the diagnosis of adrenal insufficiency. A random cortisol level <3 mcg/dL confirms the diagnosis of adrenal insufficiency.

Failure to respond adequately to an adrenal stimulation test

Inappropriately low basal cortisol levels

An unequivocal clinical response to empiric exogenous steroids

Patients with septic shock

Patients taking chronic steroids: See question 10.

Patients with HIV disease: The adrenal gland may be involved in >50% of patients infected with HIV. However, because adrenal function requires <20% of the gland to function, adrenal insufficiency in this population is uncommon (3%).

Patients with cancer: Even when cancers metastasize to the adrenal gland, adrenal dysfunction is uncommon.

High-risk postoperative patients: Patients >55 years old, patients undergoing major operations (e.g., coronary artery bypass grafting, abdominal aortic aneurysm repair, Whipple procedure), patients with multiple trauma, and postoperative patients requiring vasopressors or failing to wean from mechanical ventilation appear to be at higher risk for adrenal insufficiency.

Fluid resuscitation: Patients with septic shock typically require multiple large boluses of intravenous fluids and often vasopressors to maintain effective arterial circulation. If the patient’s blood pressure responds poorly to fluids and vasopressors, the administration of stress-dose steroids should be initiated.

Steroid dosing: Administration of hydrocortisone, 300 to 400 mg/day given intravenously in three or four divided doses with or without fludrocortisone (50 mcg enterally every day), is accepted practice.

Steroid duration: For adult patients with septic shock whose blood pressure is poorly responsive to multiple intravenous fluid boluses and vasopressor for >1 to 2 hours, the author recommends administration of stress-dose hydrocortisone (300 mg/day) for 4 days. If the patient shows rapid clinical improvement, the steroids may be stopped or tapered over 1 to 2 days. If significant hypotension recurs, steroid dosing should return to the initial dose, and a rapid taper can be undertaken after 7 days.

Yes, of course: A majority of patients with septic shock have relative adrenal insufficiency. The mortality rate for such patients is 30% to 60%. A landmark randomized control trial found an absolute mortality reduction of 10% among patients with severe sepsis or septic shock and relative adrenal insufficiency who received stress-dose steroids versus placebo. This finding is supported by a recent systematic review that concluded that prolonged low-dose steroid use reduces all-cause mortality among adult patients with septic shock. Steroid supplementation for most ICU patients with septic shock makes sense given the modest risk of a short course of low-dose steroids.

No, the data are too mixed: The Annane trial did not demonstrate across-the-board mortality reduction in the steroid group with severe sepsis and septic shock. The subsequent CORTICUS study showed no mortality benefit after hydrocortisone therapy in patients with septic shock. No mortality benefit of steroids was seen even among patients who met criteria for relative adrenal insufficiency. In addition, the steroid arm of the CORTICUS trial had significantly more superinfections. Early, smaller studies indicated a survival benefit from using steroids in septic shock. However, this benefit was not seen in meta-analysis of later and larger studies. The Surviving Sepsis Campaign International Guidelines recommend using steroids for patients with septic shock unresponsive to fluids and vasopressors. This is largely a consensus-based recommendation rather than an evidence-based conclusion. The scientific evidence to support even this limited use of steroids in patients with septic shock is modest.