Adrenal insufficiency in critically ill patients with human immunodeficiency virus

Objective: The adrenal gland is the endocrine organ most commonly involved in patients infected with human immunodeficiency virus (HIV). It is important to recognize patients with adrenal insufficiency, because this disorder may be fatal if untreated. The incidence of adrenal insufficiency in critically ill HIV-infected patients is unclear, partly because different criteria are used to diagnose adrenal insufficiency. To help clarify the incidence of adrenal insufficiency in HIV-infected critically ill patients, we compared the incidence based on the stress cortisol concentration, low-dose corticotropin stimulation test, and high-dose corticotropin stimulation test. Setting: Medical intensive care unit. Patients: Twenty-eight critically ill patients with HIV, mean age 43 +/- 9 years; 20 were male; 54% died. Interventions. We performed a 1-mug (low-dose adrenocorticotropic hormone; LD-ACTH) and 249 mug (high-dose; HD-ACTH) corticotropin stimulation test in HIV-infected critically ill patients not receiving corticosteroids. According to results of the stress serum cortisol concentration and LD-ACTH and HD-ACTH tests, patients were classified as having adrenal insufficiency or as normal. The results of newly revised diagnostic criteria for adrenal insufficiency (i.e., stress cortisol concentration and cortisol response to LD-ACTH <25 mug/dL) were compared with the traditional criteria (i.e., stress cortisol level and response to HD-ACTH <18 mug/dl). In addition, the associations between adrenal insufficiency and the CD4 count, human cytomegalovirus antigenemia, and other risk factors for adrenal insufficiency were determined. Measurements and Main Results: When we used a stress cortisol concentration <18 mug/dL as the sole diagnostic threshold for diagnosis of adrenal insufficiency, 50% (14 of 28) of patients had adrenal insufficiency. The incidence was 75% (21 of 28) when we used a diagnostic threshold of <25 mug/dL for the stress cortisol concentration, When we used both the stress cordsol concentration and LD-ACTH test with a diagnostic cortisol threshold concentration <18 mug/dL, 21% (six of 28) had adrenal insufficiency. The incidence was 46% (13 of 28) when we used a cortisol diagnostic threshold concentration of <25 mug/dL. When we used both the stress cortisol concentration and the HD-ACTH stimulation test with a diagnostic cortisol threshold of <18 mug/dL, 7% (two of 28) had adrenal insufficiency. The incidence of adrenal insufficiency was 21% (six of 28) with a cortisol diagnostic threshold value of <25 mug/dL. Human cytomegalovirus antigenemia was the only variable assessed that was associated with adrenal insufficiency. Conclusion: There is a high incidence of adrenal insufficiency in critically ill HIV-infected patients that varies with the criteria used to diagnose adrenal insufficiency. The LD-ACTH stimulation test is more sensitive than the high-dose test for diagnosing adrenal insufficiency in this population. Because of the high incidence of inadequate adrenal function, all critically ill HIV-infected patients should undergo an evaluation for adrenal insufficiency with the stress cortisol concentration and LD-ACTH stimulation tests.