Introduction

Corticosteroids exert effects on virtually every organ and tissue and are unquestionably essential for maintaining homeostasis. Cortisol plays a vital role in the maintenance of normal vascular tone, endothelial integrity and the regulation of fluid movements within the extravascular compartment. Cortisol also potentiates the vasoconstrictor effects of catecholamines. Additionally, stimulation of the hypothalamic-pituitary axis resulting in elevated levels of cortisol is one of the most important hormonal reactions to severe illness, trauma, sepsis, anaesthesia and surgery. However, the usefulness of exogenous corticosteroid administration in patients with septic shock is not as straightforward.

The use of corticosteroids in septic shock has been a subject of continued debate. The controversy entered a quiet period following the publication of two meta-analyses in 1995. Each study concluded that glucocorticoids were not beneficial in sepsis and septic shock. The Cronin study actually suggested glucocorticoids might be harmful. In the past few years, the interest in steroids in the management of septic shock has been renewed. Several trials using physiological doses for longer duration and later in the course of sepsis have shown improvement in reversal of shock and encouraging trends in 28-day mortality.

A number of theories have been proposed to explain why a low dose of glucocorticoids administered for a longer duration might be beneficial to patients in later stages of septic shock, including increased catecholamine stimulation of myocardial contractility, increased vasomotor catecholamine response, prevention of desensitisation of beta receptors and increased alpha-adrenergic receptor density. However, the theory that has received the most attention is the concept of relative adrenal insufficiency.

Absolute adrenal insufficiency in critically ill patients is rare; however, a number of investigators have detected a difference in basal plasma cortisol levels and diminished responses to corticotropin-releasing hormone in patients who died of septic shock as compared to those who survived. Annane et al., conducted a prospective clinical trial to characterise the cortisol level and cortical response to corticotropin in 189 patients with septic shock. They found an elevated basal cortisol level (>938 nmol/l) and a poor response to corticotropin (<248 nmol/l elevation at 30 or 60 minutes) to be a predictor of the poorest survival in patients with sepsis.

Human Studies

Several recent prospective, randomised placebo-controlled trials have evaluated the use of low doses of steroids in 'refractory' septic shock. In the Bollaert study, septic patients requiring vasopressor therapy (dopamine >10 µg/kg/min or any dose of either epinephrine or norepinephrine) for more than 48 hours were randomised to receive either hydrocortisone (100 mg i.v. (1.4 mg/kg) q8h for 5 days) or placebo, with tapering over 6 days if shock reversal was achieved. They found a significant reduction in the time it took to reverse shock, and a trend towards decreased 28-day mortality. The treatment effect was present irrespective of hypothalamic-pituitary axis function as measured by baseline cortisol levels and after ACTH stimulation.

Briegel et al. enrolled septic shock patients with high output circulatory failure requiring vasopressor therapy (dopamine >5 µg/kg/min or any dose of either epinephrine or norepinephrine) for no more than 72 hours. Patients were randomised to receive either a placebo or hydrocortisone (100 mg i.v. bolus, followed by a continuous infusion of 0.18 mg/kg/hr) until septic shock resolved. The length of time for which vasopressor support was required was significantly reduced in the hydrocortisone group (2 days vs. 7 days). There was a trend towards earlier reversal of organ dysfunction. There was not, however, a mortality difference between the two groups.

Annane et al. demonstrated a relative risk reduction of almost 30% in patients with septic shock treated with low-dose corticosteroid and mineralocorticoid support. 300 adult patients in septic shock were randomised to receive either placebo or hydrocortisone (50 mg i.v. bolus every 6 hours) and fludrocortisone (50 µg tablet daily) for 7 days after undergoing a short corticotropin test. The survival benefit of combination therapy was significant only in patients with blunted response to ACTH stimulation.

Veterinary Studies

Information about the incidence of relative adrenal insufficiency and the potential therapeutic value of 'stress' dose corticosteroids in critically ill veterinary patients is very limited. Prittie et al. found no evidence of adrenal insufficiency in 20 critically ill dogs that were evaluated with basal plasma cortisol, response to ACTH stimulation and endogenous ACTH levels. The study was not limited to patients with vasopressor-dependent septic shock. In a similar study of critically ill cats, the investigators found that critically ill cats had higher basal cortisol levels than the control group. Basal cortisol, ACTH-stimulated cortisol and delta cortisol levels did not differ significantly between cats that survived and cats that died, or between the septic and non-septic cats. However, critically ill cats with neoplasia had lower delta cortisol levels and were more likely to die than other cats in the study population. Burkitt et al. evaluated the response to ACTH stimulation in 14 dogs with sepsis. A decreased response to ACTH stimulation was found in 40% of the dogs tested. Survival in the group of non-responders was 40% compared to 100% responders.

Diagnosis

There are several controversies surrounding the diagnosis of relative adrenal insufficiency in human patients and identification of the patient population most likely to benefit from administration of 'stress' dose corticosteroids. Reference ranges for adrenal function testing in healthy humans have been well established. However, there remains a lack of consensus regarding the appropriate response to ACTH stimulation in critically ill patients. Some investigators advocate the use of an absolute ACTH-stimulated cortisol concentration as the sole criterion for the documentation of normal or abnormal adrenal function. While others have advocated the use of change of cortisol in response to ACTH ([DELTA]max). A study by Hamrahian et al. suggested that serum free cortisol levels are a more accurate indicator of adrenal function. Currently, assays for serum cortisol measure the total hormone concentration (serum free cortisol plus the protein-bound fraction of cortisol). Free cortisol is responsible for the physiological function of the hormone. Because a large percentage of circulating cortisol in the serum is bound to proteins, it is reasonable to suggest that alterations in the binding proteins could affect measured concentrations of serum total cortisol and, thus, the interpretation of tests used to assess adrenal function. Unfortunately, the measurement of serum free cortisol levels is currently very difficult to obtain. Because of these unanswered questions, some authors have suggested that relative adrenal insufficiency be defined as rapid clinical and haemodynamic improvement in catecholamine-dependent patients after the administration of a 'stress' dose of corticosteroid.

Treatment

Guidelines established by the Surviving Sepsis Campaign recommend intravenous corticosteroids (hydrocortisone 200-300 mg/day) for patients in septic shock who, despite fluid replacement, require vasopressor therapy to maintain adequate blood pressure regardless of adrenal function. The surviving sepsis guidelines recommend low doses of corticosteroids for 'refractory' shock but not during severe sepsis without shock or mild shock, and the use of short-course, high-dose steroids is strongly discouraged.

Conclusion

Corticosteroids may have a role in the management of the most severely ill population, but the current evidence does not support their use in all patients with sepsis, and administration of steroids should not be a substitute for a careful search for other causes of unresponsive hypotension.

References

1.  Annane D, Sebille V, et al. A 3-level prognostic classification in septic shock based on cortisol levels and cortisol response to corticotropin. Journal of the American Medical Association 2000; 283(8): 1038-1045.

2.  Annane D, Sebille V, et al. Effect of treatment with low doses of hydrocortisone and fludrocortisone on mortality in patients with septic shock. Journal of the American Medical Association 2002; 288(7): 862-871.

3.  Bollaert PE, Charpentier C, et al. Reversal of late septic shock with supraphysiologic doses of hydrocortisone. Critical Care Medicine 1998; 26(4): 645-650.

4.  Briegel J, Forst H, et al. Stress doses of hydrocortisone reversed hyperdynamic septic shock: a prospective, randomized, double-blind, single-center study. Critical Care Medicine 1999; 27(4): 723-731.

5.  Burkitt JM, Haskins SC, Nelson RW. Relative adrenal insufficiency in dogs with septic systemic inflammatory response syndrome (SIRS). In: Proceedings of the 11th International Veterinary Emergency and Critical Care Symposium 2005, p. 1035.

6.  Cronin L, Cook DJ, et al. Corticosteroid treatment for sepsis: a critical appraisal and meta-analysis of the literature. Critical Care Medicine 1995; 23: 1430-1439.

7.  Hamrahian AH, et al. Measurements of Serum Free Cortisol in Critically Ill Patients. The New England Jounral of Medicine 2004; 350(16): 1629-1638.

8.  Prittie JE, Barton LJ, et al. Pituitary ACTH and adrenocortical secretion in critically ill dogs. Journal of the American Veterinary Medical Association 2002; 220(5): 615-619.