Relative adrenal insufficiency (Proceedings)

Relative adrenal insufficiency (RAI) has been recognized with increasing frequency in critically ill people, particularly in association with sepsis. The specific mechanism for the development of RAI is unclear, and likely reflects multiple factors, including elevated cytokine levels and other inflammatory mediators. Inflammatory cytokines IL-6, TNF or corticostatin peptides produced by immune cells (mostly monocytes and macrophages) may interfere with HPA axis function. Although glucocorticoids maybe produced in adequate amounts during critically illness, this may not be an adequate amount for the severity of disease present (the severity of stress on the patient and adequate response) or may not be able to exert effects on receptors or target sites. Cytokines may cause glucocorticoid receptor dysfunction. Sepsis and systemic inflammatory response syndrome (SIRS) are the most common underlying causes of RAI, with reports of its occurrence in 24-77% of people with sepsis. In critically ill people, treatment for RAI consists of exogenous low dose glucocorticoid therapy, resulting in improvement in hemodynamic stability as well as reduced morbidity and mortality. Early treatment with low doses of glucocorticoids is also routinely employed in people considered at high risk for RAI.

The investigation of an animal with suspected RAI requires the patient to have a serious illness, such as sepsis, systemic inflammatory response syndrome (eg. necrotizing pancreatitis), severe trauma, or neoplasia. If these animals are unresponsive to fluid therapy, hypotensive, and require vasopressors, one should investigate them for RAI. These patients may still have some adrenal function left, but it is not an adequate response for the severity of the illness present. Supraphysioloc doses of steroids may benefit these patients and allow them to be weaned from vasopressors. This is not a recommendation for indiscriminate use of steroids, and testing should be performed to justify the use of steroids in these patients.

Cortisol levels in critically ill people are highly variable. Several human studies have supported that patients with basal cortisol levels of ≤15 μg/dl appear to benefit from cortisol replacement, while patients with basal cortisol levels >34 μg/dl are unlikely to have RAI. An ACTH stimulation test may also be used to differentiate critically ill people with RAI, with a change of < 9 μg/dl from baseline (delta cortisol) supporting a diagnosis of acquired RAI.

Guidelines published as part of the Surviving Sepsis Campaign in the human literature recommend the administration of hydrocortisone (200 – 300 mg/day) for seven consecutive days in three or four divided doses or by continuous infusion for people with septic shock, who despite adequate fluid replacement require vasopressor therapy to maintain adequate blood pressure. These non-responders were identified using an ACTH stimulation test, with a stimulate cortisol change (delta cortisol) of < 9 μg/dl. The appropriate criteria for critically ill animals has not yet been proposed or validated. In a recent paper in the JVIM by Burkitt JM et al, septic dogs admitted to a small animal intensive care unit with a delta cortisol of ≤ 3 μg/dL were 4.1 times more likely to die than if the delta cortisol was >3 μg/dL. They concluded that a delta cortisol of ≤ 3 μg/dL after adrenocorticotropic hormone administration is associated with systemic hypotension and decreased survival in septic dogs.

Prior investigation of the potential for RAI in dogs and cats has been limited to small studies or case reports. In one study of twenty dogs with a variety of naturally-occurring illnesses, basal cortisol levels, endogenous ACTH levels and ACTH stimulated cortisol levels were evaluated at admission and then daily until death, euthanasia or discharge from the ICU. Although a loss of correlation between the measured ACTH levels and the measured cortisol concentrations was observed in a segment of the study population, the study did not identify changes consistent with relative adrenal insufficiency. The same investigators evaluated twenty cats hospitalized in an intensive care unit and compared to normal controls. Results of this study included increased basal cortisol levels as compared to the control group, but the ACTH stimulated cortisol levels did not differ between the two groups. No significant differences in measured variables were found between survivors and non-survivors, or between septic and non-septic cats. However cats with neoplasia were found to have a lower delta cortisol, and were more likely to die than other cats in the study.

Relative AI was investigated in ten cats with cytologic or histopathologic evidence of lymphoma. Based upon ultrasound examination none of the cats were found to have involvement of the adrenal glands. Nine of the cats had subnormal cortisol response to ACTH, five had high plasma ACTH concentrations, and only one had an abnormal sodium:potassium ratio. The authors concluded that many of the cats in this study had RAI. There is also a single case report of a cat with polytrauma, sepsis, and severe hypotension unresponsive to fluid therapy that responded to glucocorticoids and was suspected to have RAI based on the obtained ACTH stimulation test. The case of which will be used as an in depth case example during presentation of talk.

The decision to perform an ACTH stimulation test in this cat was based upon persistent hypotension despite fluid therapy, and vasopressor dependence. Classic signs of hypoadrenocorticism such as decreased sodium-potassium ratios are rare with acquired RAI of critical illness, likely due to concurrent fluid therapy and intact rennin angiotensin aldosterone system (RAAS). Eosinophilia is uncommon in critically ill human patients, but if present should raise the index of suspicion for RAI. In a population of high risk human surgical patients, the relative and absolute number of eosinophils were found to be significantly higher in the adrenal insufficiency group compared to the normal ACTH responders. The most important diagnostic clues in people are hemodynamic instability and vasopressor dependence, despite aggressive fluid resuscitation in patients with sepsis or significant inflammation.

Low dose corticosteriod replacement in human cases of RAI has been found to reduce vasopressor requirements and is associated with a more rapid reversal of shock. In a trial of 300 patients with septic shock, inadequate response to the corticotropin test (defined as a delta cortisol < 9 μg/dl) was obtained in 76% of patients, and reduced mortality was documented in non-responders receiving low dose corticosteroid (50 mg hydrocortisone IV q 6h) replacement compared to those receiving placebo. This study failed to document a reduction in mortality in the group of responders receiving corticosteriods when compared to placebo. In another prospective, randomized, double-blind, placebo-controlled trial of 41 patients with septic shock in which 22 patients received low dose corticosteroid (100 mg hydrocortisone IV q 8h) replacement and 19 received placebo, 15 (68%) patients receiving low dose corticosteriods and 4 (21%) patients receiving placebo achieved shock reversal based on a stabilization of blood pressure (>90 mmHg) without vasopressors. The 28 day mortality was also lower in the group receiving corticosteroids (32% vs. 63%). Reversal of shock within 7 days of therapy was a strong predictor of survival in this study of people with sepsis.

In order to diagnose RAI, other factors that could alter the pituitary adrenal axis should be excluded. The anesthetic etomidate and the use of the antifungal agent ketoconazole have been documented to inhibit the enzymes involved in cortisol synthesis. Other causes of acute adrenal insufficiency documented in people include immunosuppression from human immunodeficiency virus (HIV), tuberculosis, and fungal disease, as well as rifampin therapy. RAI usually occurs secondary to either sepsis or systemic inflammation. A positive response to low dose glucocorticoid therapy and subnormal post-stimulation cortisol concentration maybe all that is available at this time to support a diagnosis of RAI.

References

Annane D, Sebile V, Charpentier C, et al. Effect of treatment with low doses of hydrocortisone and fludrocortisone on mortality in patients with septic shock. J Am Med Assoc 2002; 288(7):862–871.

Durkan S, de Laforcade A, Rozanski L, et al. Suspect Relative Adrenal Insufficiency in a Critically Ill Cat. JVECC 17(2) 2006, pp 197-201.

Martin LG, Groman RP. Relative adrenal insufficiency in critical illness. J Vet Emerg Crit Care 2004; 14(3):149–157. Cooper MS, Stewart PM. Corticosteroid insufficiency in acutely ill patients. N Engl J Med 2003; 348(8):727–734.

Martin LG, Groman RP, Fletcher DJ, et al. Pituitary-adrenal function in dogs with acute critical illness. J Am Vet Med Assoc. 233(1) 2008, pp87-95.

Soni A, Pepper GM, Wyrwinski PM, et al. Adrenal insufficiency occurring during septic shock: incidence, outcome, and relationship to peripheral cytokine levels. Am J Med 1995; 98(3):266–271.