Insufficient cortisol production by the adrenal cortex is the hallmark of adrenal insufficiency. Primary adrenal insufficiency results when there is adrenal pathology and secondary/tertiary (or “central”) adrenal insufficiency results when there is pituitary/hypothalamic pathology affecting ACTH secretion. Decreased aldosterone reserve, as part of primary adrenal insufficiency, may occur in CI in association with cortisol insufficiency, or more rarely, alone as selective aldosterone deficiency.16 Underproduction of the weak androgen dehydroepiandrosterone, produced mainly in the zona reticularis of the cortex, or underproduction of adrenal catecholamines, produced in the adrenal medulla generally do not contribute to the clinical scenario of adrenal insufficiency.
Cortisol, the main glucocorticoid hormone, exerts its numerous systemic functions by inducing tissue-specific gene expression. During times of physiologic stress, cortisol secretion increases. Patients with insufficient cortisol production may experience signs and symptoms as mild as headache, malaise, abdominal pain, nausea (generally without vomiting), myalgias, and/or lassitude with adrenal withdrawal (associated with steroid tapering and suppression of the hypothalamic-pituitary-adrenal [HPA] axis), to signs as severe as hypotension, incapacitating fatigue, severe hypoglycemia, hyponatremia (with hyperkalemia if primary adrenal insufficiency), and acidosis with adrenal insufficiency or crisis (associated with either primary or central adrenal insufficiency). Primary adrenal insufficiency can result from autoimmune, infectious, or hemorrhagic processes within the adrenal cortex, for example, Waterhouse–Friderichsen syndrome from meningococcal infection or adrenal vein thrombosis resulting from antiphospholipid antibody syndrome. Central adrenal insufficiency typically results from overly rapid tapering of corticosteroid dosing or increased stress in a patient who was recently tapered off of corticosteroids.
During CI, a form of central adrenal insufficiency (also termed “relative adrenal insufficiency” or “adrenal exhaustion” in the medical literature) may develop due to higher systemic cortisol requirements, effects of certain medications that inhibit cortisol synthesis, such as etomidate, and/or an impaired HPA axis during CI. The incidence of this form of central adrenal insufficiency has been reported between 10% and 30% of patients admitted with septic shock.17 Note that this entity, though real and relevant, is still the subject of great interest among intensivists and changes in definitions and management are likely to change in the near future.
Adrenal insufficiency should be suspected in patients with unexplained or persistent hypotension, hypoglycemia, hyponatremia, or hyperkalemia. An absence of hyperglycemia, that is, expected as a stress response to CI may also be an indication of adrenal insufficiency.3 The diagnosis is made by cosyntropin (250 mcg) stimulation testing that includes 2 criteria to confirm normal adrenal function: (1) plasma cortisol levels should reach 18 µg/dL/h after administering recombinant ACTH; (2) plasma cortisol levels should increase by at least 9 µg/dL from baseline. If both of these criteria are met, patients should be considered to have normal adrenal function. Patients with severe adrenal insufficiency fail to meet either of these criteria. If only 1 of the 2 criteria is met, adrenal insufficiency may still be suspected based on clinical impression.17 Unfortunately, there are several legitimate caveats to this approach that account for overdiagnosis and underdiagnosis: (1) measured total cortisol levels may be falsely low due to typically suppressed levels of cortisol-bind globulin (generally paralleling suppressing of albumin), (2) inadequate normative data for different CI patient subsets, and (3) the appearance of cortisol resistance. Consequently, a free cortisol level may be helpful, but this test usually requires several days to obtain a result, rendering the test moot for emergent decision making. Thus, a clinical diagnosis is still paramount to guide management.18
Treatment of adrenal insufficiency in CI includes stress-dose corticosteroids, usually hydrocortisone at doses of 50 to 100 mg every 6 to 8 hours. In patients with suspected adrenal insufficiency, corticosteroid treatment should be empirically initiated, and may be discontinued later in the subgroup of patients that do not meet criteria after the results of cosyntropin stimulation or free cortisol testing become available. In the cases of septic shock refractory to fluid resuscitation, and in patients with acute respiratory distress syndrome, several studies demonstrate improvement in clinical outcomes with glucocorticoid therapy regardless of cosyntropin stimulation testing.18 Treatment with stress dose glucocorticoid therapy is currently recommended in these groups of patients.18 Another approach that is more clinically oriented is to provide a lower dose of hydrocortisone (50 mg q 12) to patients suspected of relative adrenal insufficiency (pressor dependence without clear reason), while deferring specific adrenal biochemical testing and then taper empirically.18
Patients with adrenal insufficiency should continue to receive steroid treatment until hemodynamically stable. As the patient improves, the corticosteroid dose should be tapered over days to weeks based on clinical judgment, taking into account initial dosing level, dosing chronicity, and setting.
The administration of fludrocortisone, a pure mineralocorticoid, in conjunction with glucocorticoid therapy in patients with septic shock and adrenal insufficiency, was suggested by the results of 1 clinical trial.19 However, a subsequent study failed to show any benefit with the use of fludrocortisone in this population.20
Several conditions that may mimic adrenal insufficiency should be considered if patients do not respond to corticosteroid therapy. Dysautonomia may occur in patients with chronic underlying neurologic disease, such as parkinsonism, which reduces sympathetic control of blood pressure and may lead to hypotension despite fluid resuscitation.21 Autonomic dysfunction may also be suspected and responsive to midodrine in patients on renal replacement therapy.22 Cortisol resistance, which may occur in preterminal patients or those with hepatic disease, may also lead to hypotension, but cortisol levels are often extremely high.23