Special Feature
Adrenal Insufficiency and Corticosteroids in the ICU
By Mark T. Gladwin, MD
During the acute stress of surgery, trauma, sepsis, pancreatitis, and other systemic cytokine-driven host responses to injury, the hypothalamic-pituitary axis (HPA) is activated. Increased corticotropin (ACTH) release from the pituitary gland promotes release of cortisol from the adrenal glands (see Figure 1A). High plasma cortisol levels are the rule in these conditions. Recent studies suggest that the adequacy of this response (i.e., the ability of the adrenal gland to increase cortisol secretion in response to corticotropin) and a relative resistance to the effects of elevated cortisol levels are significant independent predictors of mortality in critically ill patients. Furthermore, treatment with moderate doses of corticosteroids may prove beneficial in subgroups of patients in septic shock or with acute respiratory distress syndrome (ARDS) who have this "relative adrenal insufficiency."
Adrenal Dysfunction in the Critically Ill Patient
Cortisol is required to maintain vascular tone and permeability and therefore the appropriate distribution of fluids within the vascular system. In view of this, it is not surprising that adrenal failure results in hypovolemia and circulatory shock. Humans with severe adrenal insufficiency, termed adrenal crisis, may present with hypovolemic shock, characterized by reduced filling pressures and decreased cardiac output with high systemic vascular resistance, or with hyperdynamic shock, with reduced systemic vascular resistance, and a high output state. Adrenal crisis should be suspected in patients with known adrenal insufficiency who stop taking their medications secondary to acute illness or whose chronic replacement dosage is inadequate to support their acute stressed state. De novo presentation should be suspected in patients with a history of pituitary disorders or thyroiditis (Graves’ disease or Hashimoto’s thyroiditis). While the incidence of complete adrenal insufficiency in severely ill patients is probably less than 2-3%, there is increasing recognition that relative adrenal insufficiency or resistance to "normal" cortisol levels contributes to poor outcomes in these patients.1,2
Relative Adrenal Insufficiency and Cortisol Resistance: A New ICU Syndrome?
Partial destruction of the adrenal gland by infiltrative malignancies or infection (as in tuberculosis, meningococcosis, or viral or fungal infection), autoimmune adrenalitis or hemorrhage, and reduced cortisol synthesis or increased metabolism mediated by medications and cytokines can contribute to a relative hypoadrenalism that only becomes clinically apparent during severe stress. Recent data suggest that increased cytokine activity can induce tissue resistance to cortisol, possibly by altering glucocorticoid receptor function.3
Clinicians should suspect adrenal insufficiency in ICU patients with unexplained circulatory collapse, high fevers, mildly elevated blood eosinophil count, hyponatremia, hypoglycemia, non-anion-gap metabolic acidosis, and hyperkalemia. Patients with significant hypotension despite aggressive volume resuscitation and vasopressor support have a particularly high likelihood of having relative adrenal insufficiency and cortisol resistance. Recent data suggest that this subgroup of patients will have a positive clinical response to hydrocortisone replacement, as discussed below. Classic signs and symptoms of hypoadrenalism, such as nausea, vomiting, fatigue, weight loss, depression, and vitiligo, are less common and nonspecific in critically ill patients.
| Table 1 | |||
| Prognosis | Mortality (%) | Cortisol (•g/dL) | Corticotropin Stimulation (•g/dL) |
| Good | 26 | < 34 | > 9 |
| Intermediate | 67 | < 34 | < 9 |
| > 34 | > 9 | ||
| Poor | 82 | > 34 | < 9 |
Cortisol Responses to Critical Illness
Following surgical procedures or trauma, cortisol levels rise rapidly but return to baseline within one to two days. The level of elevation correlates with the invasiveness of the surgery. Laparotomy results in twice the cortisol elevation (> 50 mcg/dL) of joint, soft tissue, and head and neck surgeries. Patients with severe sepsis or multiple trauma will have elevated levels (> 25 mcg/dL) for days and cortisol values are extremely high prior to death (30-269 mcg/dL).1
Therefore, the "normal" level of cortisol in critically ill patients and the response to corticotropin stimulation tests must be redefined. It is now becoming clear that critically ill patients with high cortisol levels and a minimal response to 0.25 mg of corticotropin (synthetic ACTH: cosyntropin), defined as an increase of 9 mcg/dL or less 30-60 minutes after stimulation, have a worse prognosis. Annane and colleagues recently reported that cortisol levels and responses to corticotropin were independent predictors of mortality in 189 consecutive patients in septic shock.4 Individuals with cortisol levels less than 34 mcg/dL with a more than 9 mcg/dL increase after corticotropin had a 26% 28-day mortality rate, while those with baseline levels greater than 34 mcg/dL and an increase of 9 mcg/dL or less had an 82% 28-day mortality rate. Intermediate mortality rates were observed for those with either a high baseline cortisol level with a more than 9 mcg/dL increase with stimulation or a low baseline level with a reduced response to stimulation (see Table 1). These data suggest that patients with high levels of cortisol and a poor adrenal response to corticotropin may have a relative resistance to their high levels of cortisol and a limited adrenal cortical reserve.
Further evidence of the importance of a normal HPA axis in critically ill patients derives from the experience of the University Hospital of Glascgow in Scotland. During a two-year period between 1981-1982, its mortality rate nearly doubled to 44% coinciding with its use of continuous infusions of etomidate.5 This medication is now known to inhibit adrenal 11beta-hydroxylase, which is required to convert deoxycortisol to cortisol. This causes partial adrenal insufficiency, which was sufficient to increase mortality in patients with multi-organ failure.
Corticosteroid Therapy Revisited in Sepsis and ARDS
Multiple studies have been performed evaluating the outcome of corticosteroid treatment of septic shock. Most of these studies involved relatively higher doses of steroids, early treatment, and a brief duration of treatment. A recent meta-analysis of 49 published trials documented no overall beneficial effect.6 Another meta-analysis evaluated nine randomized, controlled trials of corticosteroids in sepsis and septic shock and similarly found no benefit. The studies with the highest methodological quality scores actually showed a trend toward a harmful effect (greatest for those in shock).7 In contrast to these data, a recent small (41 patients), prospective, randomized, double-blind, placebo-controlled trial of hydrocortisone therapy (100 mg q 8 hrs for 5 days; see Table 2) for ICU patients in septic shock (requiring catecholamines for > 48 hours) found a significant reversal of shock and a 31% reduction in absolute mortality (32% vs 63%).8
| Table 2 | |
| Indications | Therapy |
| • Septic shock requiring vasopressors | |
| • Cortisol levels increase < 9 • g/dL 30-60 minutes after 0.25 mg corticotropin (ACTH) | • Hydrocortisone 100 mg three times daily for five days |
| • Severe ARDS on mechanical ventilation for seven days with no improvement | Methylprednisolone: |
| • No evidence of infection | • Loading dose: 2 mg/kg |
| • 2 mg/kg/day divided q 6 hrs from day 1-14 | |
| • 1 mg/kg/day divided q 6 hrs from day 15-21 | |
| • 0.5 mg/kg/day divided q 6 hrs from day 22-28 | |
| • 0.25 mg/kg/day divided q 6 hrs from day 29, 30 | |
| • 0.125 mg/kg/day divided q 6 hrs from day 31, 32 | |
| • Surveillance bronchoscopy to rule out pneumonia every week | |
A large ongoing clinical trial in France evaluating corticosteroids in patients with septic shock and a less than 9 mcg/dL increase in cortisol after corticotropin stimulation is reportedly showing similar salutary effects (preliminary data presented at the 5th World Congress on Trauma, Shock, Inflammation and Sepsis, March 3, 2000, Munich, Germany, by D. Annane). The observed reduction in mortality in the setting of vasopressor-dependent septic shock suggests that this subgroup of patient may be suffering from relative adrenal insufficiency, and that stress doses of hydrocortisone may compensate for tissue resistance to cortisol.
In patients with ARDS in a late stage, similar dramatic effects on mortality have been reported by Meduri and colleagues. Patients who had ARDS and a lung injury score of 2.5 or greater were randomized to receive methylprednisolone or placebo for 32 days (see Table 2). All patients received surveillance bronchoscopy on study day 5 and weekly thereafter to exclude pneumonia. Treatment with corticosteroids resulted in a reduction in lung injury score, an improved ratio of arterial PO2 to inspired oxygen fraction (FiO2), reduced multi-organ dysfunction, increased rates of extubation, and improved ICU and hospital mortality.9 Surveillance bronchoscopy was critical with this treatment because pneumonia frequently occurred in the absence of fever. Whether this therapy improves outcome by reducing inflammation and fibrosis in the lung or is attributable to the treatment of undiagnosed relative adrenal insufficiency remains unanswered. The ARDS Network is currently evaluating corticosteroid therapy for late ARDS in a large randomized trial.
How then can we explain these new reports documenting a significant mortality benefit from corticosteroids for septic shock and ARDS when so many previous studies showed no benefit? Meduri et al argue that these new studies differ in the timing, dosage, and duration of therapy.3 These studies involved the late delivery of stress doses of steroids during either vasopressor-dependent shock or the fibroproliferative phase of ARDS, and a prolonged treatment course sufficient to compensate for tissue resistance to cortisol and to suppress inflammation for the duration of tissue repair.
Therapy of Acute Adrenal Crisis
Adrenal crisis should be managed with aggressive volume resuscitation, with dextrose if hypoglycemia is present, and with cortisol replacement. Maximal cortisol production is estimated to be 200-400 mg over 24 hours. Therefore, an appropriate "stress dosage" of hydrocortisone is 100 mg every 6-8 hours.2 Hydrocortisone is an ideal corticosteroid for adrenal crisis because it also has considerable mineralocorticoid activity. One problem with hydrocortisone is that the radioimmunoassay or HPLC cortisol assays, used to measure cortisol during the synthetic ACTH stimulation test, will be falsely elevated after hydrocortisone is given. If diagnostic tests are to be performed, dexamethasone, which does not interfere with the cortisol assays, should be given at a dose of 2-4 mg every 12 hours.
Conclusions
It is likely that many critically ill patients, particularly those with multiple organ dysfunction and shock, have a syndrome of relative adrenal insufficiency characterized by a target organ resistance to the action of cortisol. The classic patient will be hypotensive on vasopressors, with an elevated cortisol level and diminished response to corticotropin stimulation (£ 9 mcg/dL cortisol). Recent studies suggest that stress-dose hydrocortisone will improve clinical outcomes in this patient group and a prolonged course of methylprednisolone may have a similar salutary effect in patients with ARDS.
References
1. Lamberts SW, et al. Corticosteroid therapy in severe illness. N Engl J Med 1997;337:1285-1292.
2. Malchoff CD, Carey RM. Adrenal insufficiency. Curr Ther Endocrinol Metab 1997;6:142-147.
3. Meduri GU, Kanangat S. Glucocorticoid treatment of sepsis and acute respiratory distress syndrome: Time for a critical reappraisal. Crit Care Med 1998;26:630-633.
4. Annane D, et al. A 3-level prognostic classification in septic shock based on cortisol levels and cortisol response to corticotropin. JAMA 2000;283:1038-1045.
5. Ledingham IM, Watt I. Influence of sedation on mortality in critically ill multiple trauma patients. Lancet 1983;1:1270.
6. Lefering R, Neugebauer EA. Steroid controversy in sepsis and septic shock: A meta-analysis [see comments]. Crit Care Med 1995;23:1294-1303.
7. Cronin L, et al. Corticosteroid treatment for sepsis: A critical appraisal and meta-analysis of the literature [see comments]. Crit Care Med 1995;23:1430-1439.
8. Bollaert PE, et al. Reversal of late septic shock with supraphysiologic doses of hydrocortisone [see comments]. Crit Care Med 1998;26:645-650.
9. Meduri GU, et al. Effect of prolonged methylprednisolone therapy in unresolving acute respiratory distress syndrome: A randomized controlled trial [see comments]. JAMA 1998;280:159-165.
Attention CRC Subscribers
CRC question no. 15 in the May 2000 issue of Critical Care Alert should have been question no. 14. We regret any confusion this may have caused.
Administration of which of the following agents does not interfere with cortisol assays?
a. Hydrocortisone
b. Dexamethasone
c. Triamcinolone
d. Methylprednisolone
Subscribe Now for Access
You have reached your article limit for the month. We hope you found our articles both enjoyable and insightful. For information on new subscriptions, product trials, alternative billing arrangements or group and site discounts please call 800-688-2421. We look forward to having you as a long-term member of the Relias Media community.