Contents
Overview
The adrenal stress response is essential for haemodynamic stability, immune modulation, and metabolic adaptation during critical illness. Adrenal insufficiency (AI) in the ICU may be classical (primary or secondary) or relative — an inadequate cortisol response for the degree of physiological stress. Recognition and treatment of AI can be life-saving.
Pathophysiology
Under physiological stress, the hypothalamic-pituitary-adrenal (HPA) axis is activated: hypothalamic CRH stimulates pituitary ACTH release, which drives adrenal cortisol synthesis from cholesterol. Cortisol exerts glucocorticoid effects (gluconeogenesis, anti-inflammation, vascular sensitivity to catecholamines) and some mineralocorticoid effect.
Critical illness-related corticosteroid insufficiency (CIRCI) describes an inadequate cortisol response relative to the degree of stress. Proposed mechanisms include:
- Relative adrenal insufficiency — reduced adrenal synthetic capacity from ischaemia, haemorrhage, or drug effects
- HPA axis suppression — direct pituitary or hypothalamic dysfunction, seen in sepsis and TBI
- Peripheral corticosteroid resistance — glucocorticoid receptor downregulation or post-receptor signalling failure
- Reduced corticosteroid-binding globulin (CBG) — in inflammation, CBG falls, reducing total cortisol without necessarily reflecting free (active) cortisol
The interplay of these mechanisms makes diagnosis in critical illness unreliable using standard biochemical tests.
Aetiology
Primary adrenal insufficiency (Addison's disease):
- Autoimmune — most common cause in the UK (80%); associated with other autoimmune conditions (thyroid disease, type 1 diabetes)
- Haemorrhage — Waterhouse-Friderichsen syndrome (bilateral adrenal haemorrhage in meningococcal septicaemia); anticoagulant therapy; trauma
- Metastatic disease — bilateral adrenal metastases (lung, breast, melanoma)
- Infection — TB, fungal (histoplasmosis, cryptococcus), CMV in HIV
Secondary adrenal insufficiency:
- Pituitary disease — tumour, apoplexy, surgery, radiation
- Prolonged exogenous corticosteroid therapy — even a few weeks of moderate-dose steroids can suppress the HPA axis; inadequate perioperative stress dosing precipitates crisis
Drug-induced:
- Etomidate — a single induction dose inhibits 11-beta-hydroxylase, suppressing cortisol synthesis for up to 24 hours. Its use for rapid sequence induction in septic patients remains controversial because of this.
- Ketoconazole, fluconazole — inhibit adrenal steroidogenesis
- Rifampicin — induces cortisol metabolism (relative insufficiency)
Diagnosis
Clinical suspicion is the most important trigger. Consider AI in:
- Refractory vasopressor-dependent shock despite adequate fluid resuscitation
- Unexplained hypoglycaemia
- Hyponatraemia with hyperkalaemia (primary AI)
- Eosinophilia
- History of steroid use, pituitary disease, or known autoimmune conditions
Biochemical testing:
A random serum cortisol below 276 nmol/L in a severely ill patient is suggestive of AI. Interpretation of cortisol in critical illness is complicated by altered CBG and variable assay calibration.
The short Synacthen test (SST) measures cortisol at 0 and 30–60 minutes after 250 micrograms of synthetic ACTH IV. A peak cortisol below 500 nmol/L is considered inadequate. The SST assesses adrenal reserve, not basal axis function; it may miss secondary AI. In critical illness, a delta cortisol response below 250 nmol/L has been proposed as the threshold for CIRCI, though this is not universally accepted.
ACTH level differentiates primary (high ACTH) from secondary (low or inappropriately normal ACTH) AI.
Management
Acute Addisonian Crisis
This is a medical emergency presenting with profound hypotension, vomiting, and altered consciousness.
- Hydrocortisone 100 mg IV bolus immediately, then 200 mg/24 hours by continuous infusion (or 50–100 mg 6-hourly)
- Intravenous fluid resuscitation with 0.9% sodium chloride — may need several litres
- 10% glucose if hypoglycaemic
- Identify and treat the precipitant
Fludrocortisone is not required acutely — the high doses of hydrocortisone provide sufficient mineralocorticoid effect. It is added once the patient is on maintenance therapy.
CIRCI in Septic Shock
Hydrocortisone 200 mg/day (continuous infusion or intermittent dosing) is used in patients with septic shock that is refractory to adequate fluid and vasopressor therapy. Evidence from large trials shows faster vasopressor weaning and shorter time to shock resolution, but no consistent mortality benefit. Current Surviving Sepsis Campaign guidance recommends hydrocortisone if haemodynamic stability cannot be achieved with fluids and vasopressors.
Weaning should be guided by clinical response, particularly vasopressor requirements. Abrupt cessation can cause rebound haemodynamic instability.
Key Trials
CORTICUS (Sprung et al., NEJM 2008) — 499 patients with septic shock randomised to hydrocortisone 50 mg 6-hourly vs placebo. No difference in 28-day mortality. More rapid shock reversal in the hydrocortisone group, but higher rates of superinfection. Challenged the previously suggested mortality benefit from an earlier French trial.
ADRENAL (Venkatesh et al., NEJM 2018) — 3658 ventilated patients with septic shock randomised to hydrocortisone 200 mg/day infusion vs placebo. No difference in 90-day mortality (27.9% vs 28.8%). Hydrocortisone was associated with faster shock reversal, earlier weaning from ventilation, and shorter ICU stay. No excess infection. This is the largest trial in this area.
APROCCHSS (Annane et al., NEJM 2018) — 1241 patients randomised to hydrocortisone plus fludrocortisone vs placebo. Reduction in 90-day mortality (43% vs 49.1%). Criticism: higher baseline severity than ADRENAL, different patient population. The addition of fludrocortisone remains uncertain in clinical practice.
Viva Questions
What is CIRCI and how does it differ from classical adrenal insufficiency?
Critical illness-related corticosteroid insufficiency describes a state in which the adrenal gland and HPA axis fail to mount an adequate cortisol response for the degree of physiological stress, despite the absence of structural adrenal or pituitary pathology. It encompasses impaired corticosteroid synthesis, altered binding protein levels, tissue resistance to glucocorticoids, and HPA axis dysfunction. Classical adrenal insufficiency, by contrast, implies structural or functional loss of adrenal or pituitary capacity — from autoimmune destruction, haemorrhage, metastatic disease, or chronic exogenous steroid suppression. The distinction matters because CIRCI is reversible, responds to replacement-dose hydrocortisone, and does not necessarily require long-term steroid therapy after recovery from the acute illness. Classical AI requires lifelong replacement.
How would you investigate a patient with vasopressor-dependent septic shock for adrenal insufficiency?
The clinical picture is key: refractory vasopressor dependence, hyponatraemia, hyperkalaemia, hypoglycaemia, and eosinophilia should raise suspicion. A random cortisol should be taken before administering hydrocortisone if time permits — a value below 276 nmol/L in a severely ill patient is suggestive but not diagnostic. A short Synacthen test (250 micrograms IV) with cortisol at 0 and 30 minutes can be performed promptly; a peak below 500 nmol/L is considered inadequate. ACTH level distinguishes primary (elevated ACTH) from secondary AI. Importantly, if clinical suspicion is high and the patient is severely compromised, empirical hydrocortisone should not be delayed while awaiting test results. Blood for cortisol and ACTH should be taken first. Further investigation — adrenal imaging, pituitary MRI, or autoimmune serology — can be pursued once the patient is stabilised.
What is the evidence for corticosteroids in septic shock?
Three major trials inform current practice. CORTICUS (2008) found no mortality benefit from hydrocortisone in a general septic shock population, though shock reversed more quickly. ADRENAL (2018), the largest trial with 3658 patients, confirmed no mortality benefit from hydrocortisone 200 mg/day continuous infusion but demonstrated faster shock reversal and shorter ICU stay. APROCCHSS (2018) reported a mortality benefit from combined hydrocortisone and fludrocortisone, but enrolled a higher-severity population, making direct comparison with ADRENAL difficult. Current Surviving Sepsis Campaign guidance recommends hydrocortisone 200 mg/day in patients whose shock is not adequately controlled despite appropriate fluids and vasopressors — acknowledging the benefit in shock reversal even without a consistent mortality signal. Routine use in all septic shock patients is not supported by the evidence.