Intended for healthcare professionals

Education And Debate

Lesson of the Week: Adrenocortical failure in intensive care

BMJ 1995; 310 doi: https://doi.org/10.1136/bmj.310.6989.1253 (Published 13 May 1995) Cite this as: BMJ 1995;310:1253
  1. N F Quiney, senior registrara,
  2. M A Durkin, consultanta
  1. a Department of Anaesthesia and Intensive Care, Gloucestershire Royal Hospital, Gloucester GL1 3NN
  1. Correspondence to: Dr Quiney.
  • Accepted 13 September 1994

Diagnosis of acute adrenocortical failure in patients admitted to intensive care units is difficult as the clinical features are non-specific and may mimic those of other conditions. We report on a patient in whom the initial diagnosis of septic shock had to be amended later to acute adrenocortical failure. The criteria for laboratory diagnosis of adrenocortical failure in patients in intensive care may need to be modified as adrenocortical function is altered in patients with severe organ dysfunction.

Case report

A 45 year old woman with myasthenia gravis was admitted to intensive care breathless and disoriented. She had a fever, with a temperature of 38°C, and had an arterial pressure of 60/40 mm Hg. Her serum concentration of sodium was 131 mmol/l and of potassium was 4.4 mmol/l. A full blood count showed a haemoglobin concentration of 148 g/l and leukocytosis, with a cell count of 18.9 x 109/l. Despite fluid resuscitation with 2000 ml 0.9% saline and 1000 ml gelatin (Haemaccel) she remained hypotensive. A pulmonary artery catheter was inserted, and the haemodynamic profile showed a normal mean pulmonary artery pressure of 15 mm Hg and cardiac index of 3.8 l/min/m2 but a low systemic vascular resistance index of 971 dyne.sec/cm5/m2 and left ventricular stroke work index of 20.7 g/m2/beat.

A presumptive diagnosis of septic shock was made, and after culture samples of blood, urine, and bronchial aspiration were taken, broad spectrum antibiotics were started. A supine chest x ray film and abdominal ultrasound scan showed no evidence of focal infection. Intravenous infusions of dobutamine (5 μg/kg/min), dopamine (3 μg/kg/min), and noradrenaline (0.05 μg/kg/min) were started. The dosage of dobutamine was increased slowly to 15 μg/kg/min and of noradrenaline to 0.75 μg/kg/min to maintain the cardiac index and systemic vascular resistance within the normal ranges.

As no obvious source of infection and only a moderate fever were present a stimulation test with tetracosactrin was carried out to exclude adrenocortical failure. The baseline cortisol concentration was 222 nmol/l, increasing to 267 nmol/l 30 minutes after 250 μg of tetracosactrin had been given. These results confirmed a diagnosis of adrenocortical failure. Within 12 hours of the start of steroid replacement with hydrocortisone, the dose of noradrenaline had been reduced by 66% and the infusion of dobutamine had been stopped. Within a further 12 hours the infusion of noradrenaline had also been stopped. A haemodynamic profile performed 48 hours after the start of steroid replacement showed that the systemic vascular resistance and the left ventricular stroke work index had both returned to normal.

Clinical features of acute adrenocortical failure

  • Tachycardia and hypotension

  • Fever

  • Oliguria

  • Confusion

  • Gastrointestinal complaints

A follow up investigation showed a normally functioning pituitary and hypothalamus but raised adrenal autoantibodies. This finding confirmed a diagnosis of primary adrenocortical failure due to autoimmune destruction of the adrenal cortex.

Discussion

The incidence of adrenocortical failure in the general population is about 5/100000, but it may be more common in patients in intensive care.1 Most cases are due to autoimmune destruction of the adrenal gland.2 Adrenocortical failure may also be caused by severe bacterial and fungal infections (for example, meningococcaemia and Pseudomonas aeruginosa), anticoagulant treatment, or previous treatment with steroids or other drugs (for example, rifampicin, spironolactone, or etomidate) or be associated with a variety of other autoimmune disorders.3

The clinical signs of adrenocortical failure are non-specific and may mimic those of other conditions (box). Laboratory diagnosis of acute adrenocortical failure may be difficult in patients with multiple organ dysfunction. The classic features of hyponatraemia and hyperkalaemia may not be present (as in this case) or may be attributed to inappropriate intravenous fluid replacement. A random cortisol concentration of <280 nmol/l has been suggested as diagnostic of acute adrenal insufficiency in acute illness.4 Clayton suggested that the best method to evaluate adrenal function is the stimulation test with low dose tetracosactrin.5 A baseline cortisol concentration of <100 nmol/l with a peak concentration of <400 nmol/l after a 250 μg bolus of tetracosactrin strongly suggests adrenocortical failure.6

Patients admitted to intensive care have been found to have baseline cortisol concentrations of between 212 and 8430 nmol/l.7 Baseline cortisol concentrations alone are not therefore adequate to diagnose acute adrenocortical insufficiency in patients with severe organ dysfunction. A stimulation test with tetracosactrin should be carried out to confirm the diagnosis; cortisol concentrations should double 30 minutes after stimulation.3

Treatment with intravenous steroid replacement should be started after the diagnosis has been confirmed. If a patient's life is in danger dexamethasone should be given without waiting for the results of cortisol assays, as it provides the glucocorticoid effects (but not mineralocorticoid effects) of cortisol without interfering with further stimulation testing with tetracosactrin.4 After acute adrenocortical failure has been diagnosed hydrocortisone should replace the dexamethasone to provide both glucocorticoid and mineralocorticoid effects. The arterial pressure should improve within two to six hours of the start of steroid replacement. Baldwin suggested that if a reduction of 75% in inotropic support within 48 hours of the start of steroid replacement is possible then this is proof of a response to the steroid replacement, although other concurrent disease may mask this response.8

The low incidence of acute adrenal failure in patients who are in intensive care makes it hard to justify routine testing of all patients admitted to intensive care. Adrenocortical function should be assessed in any patient in whom a diagnosis of adrenocortical failure is considered.

Studies of patients with acute adrenocortical insufficiency show that the cardiovascular findings may be described as one of two distinct types. All patients have appreciable hypotension, but some patients have hyperdynamic shock with a high cardiac index and a low systemic vascular resistance index (box). Other patients have myocardial depression with decreased plasma volume and a raised systemic vascular resistance index. After fluid resuscitation patients with myocardial depression and a raised systemic vascular resistance index have been found to have the haemodynamic features of hyperdynamic shock with a high cardiac index and low systemic vascular resistance index. Giving glucocorticoid agents to either group results in an increase in myocardial contractility, with an increase of at least 20% in the left ventricular stroke work index.9

Differential diagnosis of hyperdynamic shock

  • Acute adrenocortical failure

  • Septic shock

  • Thyrotoxicosis

  • Arteriovenous shunt

  • Anaemia

  • Pregnancy

  • Beriberi

  • Liver failure

We thank Drs A McCrirrick and A Williams for their help in preparing this manuscript.

Practical points

  • A high index of suspicion is required in all patients admitted to intensive care to identify those patients with adrenocortical failure

  • Acute adrenocortical failure is best diagnosed with stimulation testing with tetracosactrin

  • Steroid replacement treatment should not be delayed by waiting for cortisol concentrations to be reported

  • Treatment with dexamethasone will not interfere with stimulation testing with tetracosactrin

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
View Abstract