For hospitalized adults with adrenal insufficiency, the expert recommendation is to double or triple the usual outpatient dose of glucocorticoid; however, data to support this is limited, and each patient should be assessed carefully and monitored to determine the optimal dose adjustment.
by Megan Shaw, MD, and Jennifer Best, MD, FACP, FHM
A 46-year-old woman with Addison’s disease and type II diabetes presents with one day of right leg pain, swelling, and redness. She has had mild nausea and vomiting over the past week, with an episode of diarrhea three days prior. She takes hydrocortisone 30mg in the morning and 10mg at bedtime, as well as fludrocortisone 0.2mg in the morning. She is afebrile with a pulse of 108 beats per minute. Her initial blood pressure was 74/49 mmHg, which improved to 84/45 mmHg following one liter of normal saline. She is mentating appropriately. The physical exam is significant for a large, tender area of erythema and warmth from the right ankle to mid-calf. She is admitted for cellulitis and intravenous antibiotics are initiated.
Does she require an increase in her glucocorticoid dose during her acute illness?
Adrenal insufficiency occurs in approximately 5 out of every 10,000 people and results from primary failure of the adrenal gland, or secondary impairment of the hypothalamic-pituitary-adrenal (HPA) axis, which regulates cortisol secretion.1 In developed countries, 90% of primary adrenal insufficiency (Addison’s disease) cases are due to autoimmune adrenalitis, which might occur in isolation or as part of an autoimmune polyglandular syndrome.1,2 Secondary adrenal insufficiency is most commonly the result of chronic glucocorticoid therapy, though lesions involving the hypothalamus or pituitary gland might be implicated.2,3
In a healthy individual, cortisol is secreted in a diurnal pattern from the adrenal glands under the control of corticotropin (ACTH) produced by the pituitary gland and corticotropin-releasing hormone (CRH) produced by the hypothalamus (see Figure 1, p. 19). In the normal state, during periods of such systemic stress as illness, trauma, burns, or surgery, cortisol production increases to a degree roughly proportional to the degree of illness (as much as sixfold).4,5 Patients with adrenal insufficiency are unable to mount an appropriate cortisol response and, therefore, are at risk for adrenal crisis—a life-threatening condition characterized by hypotension and hypovolemic shock.
Although recommendations for high-dose intravenous steroids in adrenally insufficient patients who are critically ill or undergoing surgery have been extensively discussed in the literature, there are relatively few data regarding the appropriate management of adrenal insufficiency in patients hospitalized for noncritical illness.
Several recent studies have investigated the patient characteristics, situations, and conditions most likely to provoke adrenal crisis in order to establish guidelines dictating the use of supra-physiologic steroid dosing.
Studies have estimated the prevalence of adrenal crisis in patients with underlying insufficiency at 3.3 to 6.3 events per 100 patient years, with 42% of patients reporting at least one crisis.2,5,6 A recent survey of 982 patients with Addison’s disease in the United Kingdom reported an 8% annual frequency.7
A retrospective Japanese study reviewed the medical charts of 137 adult patients receiving steroid replacement for established primary or secondary adrenal insufficiency. The authors noted a significant positive association between adrenal crisis and long-term steroid replacement (>4 years), concomitant mental disorder, and sex hormone deficiency. A combination of any of these factors further increased the risk.8
A more recent survey of 444 patients ages 17-81 assessed independent risk factors for adrenal crisis in the setting of primary (N=254) or secondary (N=190) adrenal insufficiency. The incidence of crisis was higher in primary versus secondary insufficiency. In patients with primary insufficiency, concomitant, non-endocrine disorders increased the risk of adrenal crisis, whereas diabetes insipidus and female gender increased risk in patients with secondary insufficiency. This same study also investigated events leading to a crisis and found gastrointestinal infection to be the most frequent factor, followed by other infectious or febrile illnesses. Overall, infection comprised 45% of all identified triggers.6
A similar study conducted by White and Arlt evaluated 841 Addison’s patients in the United Kingdom, Canada, Australia, and New Zealand.7 Again, gastrointestinal illness was the most common provoking factor, responsible for 56% of all reported crises. Flulike illness followed at 11%, followed by infections and surgical procedures at 6% each. This study found a higher risk of crisis in patients with diabetes (type I or II), premature ovarian failure, and asthma; the presence of multiple comorbidities further increased risk.
Medications. Glucocorticoid therapy is known to suppress the HPA axis. Although it was once believed that the duration and dose of therapy correlated directly with the degree of HPA suppression, more recent studies have failed to find any definite relationship.9 Patients taking the equivalent of 5mg of prednisone per day should continue to have an intact HPA axis, as this dose mimics physiologic secretion of cortisol in a healthy individual.3
However, the dose and duration of therapy at which suppression occurs is highly variable between patients. In general, patients on 7.5mg of prednisone or more per day for at least three weeks should be considered to be suppressed.3,9 Additionally, progesterone derivatives (i.e., megestrol) have glucocorticoid activity and might suppress HPA function. Other medications that might have related effects are those that inhibit enzymes involved in cortisol synthesis; ketoconazole and etomidate are common examples. Rifampin and several classes of anti-epileptic drugs induce enzymes, which increase hepatic metabolism of cortisol (see Table 1, p. 18).
Hyperthyroidism. Thyroid hormone is involved in the metabolism of cortisol, thus an increase in T4 correlates with lower levels. Adrenal insufficiency and thyroid disease might coexist within the autoimmune polyglandular syndrome. Initiation or uptitration of thyroid replacement should be avoided if acute adrenal insufficiency is suspected, as this might provoke an adrenal crisis. Conversely, any patient with adrenal insufficiency who has uncontrolled hyperthyroidism should receive two to three times their usual glucocorticoid replacement.1,2
Pregnancy. Levels of cortisol-binding globulin increase throughout pregnancy. In women with intact adrenal function, free cortisol levels also rise during the third trimester. Therefore, glucocorticoid replacement should be increased by 50% during the last three months of pregnancy.2
Acute illness. The cortisol response to stress is highly variable and dependent on a number of factors, including age, underlying health, and genetics. In general, most experts recommend doubling or tripling the daily replacement dose during mild to moderate illness until recovery (often referred to as the “sick-day rules”). What constitutes “recovery” is not clearly defined. When oral intake is compromised, as with vomiting or diarrhea, parenteral administration of steroids is recommended.1-5,9,10 Patients with adrenal insufficiency should be provided an emergency injection kit to use and further counseled to seek medical attention. Injection kits typically consist of 100mg of hydrocortisone or 4mg of dexamethasone, although other glucocorticoids may be used (see Table 2, above left, for conversions).
Limited data are available to support guidelines for glucocorticoid adjustment during acute, non-critical illness. Published guidelines vary both in illness categorization and category-specific recommendations (see Table 3, below).
Coursin and Wood devised a set of guidelines based on a literature review and consultation with experts, categorizing medical illness as minor, moderate, severe, and critical (see Table 3).3 For noncritical illness, they recommended continuation of standard replacement therapy with an additional, once-daily dose, which varied according to illness severity.
Cooper and Stewart conducted a similar review, basing their guidelines on a categorization of mild, moderate, severe/critical, or septic shock. These guidelines recommended a total daily dose of glucocorticoid supplementation, rather than an addition of a single dose to current therapy. They also stated that the least severe category of illness (defined as mild) did not require any change to a patient’s regular therapy.4
Jung et al classified illness as minimal, minor, moderate, severe, and critical.9 Under these guidelines, supplemental therapy was not advised for minimal (nonfebrile) illness. Moderate illness, including cellulitis, warranted a doubling or tripling of the outpatient dose until recovery, which was consistent with prior expert recommendation. More severe illness warranted intravenous administration of steroids.
The patient had a mild case of cellulitis, classified by most experts as moderate illness, which responded well to vancomycin. Her outpatient glucocorticoid dose was doubled on admission and administered orally for the duration of her hospitalization, as she had no further episodes of vomiting or diarrhea.
Review of the patient’s records from prior hospitalizations and ambulatory visits revealed that her systolic blood pressure typically ran in the 80 mmHg to 100 mmHg range. Following initial volume resuscitation, her systolic blood pressure remained in the 90-100 mmHg range.
She was discharged home in stable condition, with instructions to complete a course of oral trimethoprim/sulfamethoxazole, resume her baseline dose of hydrocortisone the day after discharge, and follow up with her PCP for further monitoring and adjustment of her adrenal replacement therapy.
For adults with adrenal insufficiency hospitalized with noncritical, nonsurgical illness, the expert recommendation is to double or triple the usual outpatient dose of glucocorticoid; however, data to support this is limited, and each patient should be assessed carefully and monitored to determine the optimal dose adjustment. TH
Dr. Shaw is a resident in the Department of Medicine at the University of Washington School of Medicine in Seattle. Dr. Best is an assistant professor of medicine in the division of general internal medicine, University of Washington School of Medicine.
The Hospitalist newsmagazine reports on issues and trends in hospital medicine. The Hospitalist reaches more than 25,000 hospitalists, physician assistants, nurse practitioners, residents, and medical administrators interested in the practice and business of hospital medicine.