Adrenocorticotropic Hormone (Cosyntropin) Stimulation Test

Etiology and Epidemiology

AI can be primary, secondary, or tertiary depending on etiology. Primary AI is also called Addison disease (AD) named after Thomas Addison, who described it first in 1855. Primary AI (AD) may be caused by autoimmunity, genetic factors, infection, hemorrhage, metastasis, or drugs.[1] AD is a very rare disorder and can affect 1 in 100,000 people. The most common cause of AD is autoimmunity in developed countries.[2]

Central AI could be secondary to decreased adrenocorticotropin (ACTH) secretion from the pituitary (secondary AI) or decreased corticotropin-releasing hormone from the hypothalamus (tertiary AI). Exogenous steroids cause decreased ACTH secretion and thus secondary AI. Central AI is much more common than primary AI, and prevalence is estimated to be 150 to 280 per million.[3]

The term relative AI was introduced in 1991 by Rothwell et al. for decreased responsiveness of adrenals after injecting 250 mcg corticotropin in critically ill patients.[4] The term was later changed to Critical Illness Related Corticosteroid Deficiency in 2008 by Marik et al.[5]

Pathophysiology

The hypothalamus secretes corticotropin-releasing hormone (CRH), which stimulates the release of ACTH from the pituitary. CRH, in turn, is stimulated by several factors and hormones like ANS (atrial natriuretic peptide), vasopressin, epinephrine, and norepinephrine. ACTH then stimulates the adrenal cortex to release cortisol, which in turn provides negative feedback to the pituitary decreasing ACTH secretion. This maintains cardiovascular function, among other things, in critically ill patients. This is a dynamic process and is constantly influenced by several factors. Adrenal insufficiency is an inability to produce sufficient cortisol at baseline and/or under stress.

The cortisol physiology in critically ill patients is slightly different and warrants some discussion. There is an initial rise of both ACTH and cortisol in response to critical illness. After the initial phase, although cortisol remains high, ACTH levels may decline. It is hypothesized that factors other than ACTH like the adrenergic system, cytokines, adipokines, and neuropeptides may be responsible for sustained cortisol rise although ACTH levels are low. Also, there may be a decrease in cortisol metabolism, contributing to higher levels. There is some evidence that there is a higher ratio of cortisol to cortisone in critically ill or trauma patients. This reflects the activity of 11-beta-hydroxysteroid in these patients.[6]

Specimen Requirements and Procedure

Administration of a supraphysiological dose of synthetic ACTH or cosyntropin stimulates the pituitary and thereby a release of cortisol from the adrenal cortex, as long as the adrenal cortex has a functional reserve. CST involves checking cortisol levels at baseline, 30 minutes, and 60 minutes after IV or IM administration of 250 mcg cosyntropin.

A negative (normal) result occurs if the cortisol level is stimulated to greater than 500 to 600 nmol/L.[7] A positive (abnormal or subnormal) result can be from primary or central AI and can be further evaluated by ACTH levels. ACTH level is expected to be high with primary and low with central AI.

AI can be diagnosed if cortisol is unequivocally low, but intermediate or higher levels need further evaluation by CST. There is no universally agreed-upon definition or criteria for the diagnosis of adrenal insufficiency (AI) in critically ill patients.[6]

Testing Procedures

CST is a standard test for the evaluation of AI. A CST test can be performed with a high dose of cosyntropin or a low dose. High dose CST (250 mcg) is considered a gold standard to diagnose primary AI. 1 mcg CST or low dose CST can be considered in the evaluation of secondary or tertiary AI.[8] It is performed similarly to high dose CST and utilizes the same cut-offs for diagnosis, but utilizes 1 mcg cosyntropin instead of 250 mcg.[6] Results should be cautiously interpreted, especially for secondary AI.

Baseline or morning ACTH (adrenocorticotropin hormone) levels help to differentiate primary from secondary AI.[8] 

Interfering Factors

CST can be misleading in some cases and warrant careful interpretation. Unless results are straightforward, evaluation should be done by an Endocrinologist or provider with expertise in adrenal function.

CST may be falsely negative or normal in mild disease or disease of recent onset.  In mild disease, there may be sufficient adrenal functional cortex to maintain adrenal reserve, which can respond adequately to the external supraphysiologic dose of cosyntropin. Similarly, in the disease of recent onset, there may not be enough time for adrenals to lose complete function and they will respond to the supraphysiological exogenous stimulus of cosyntropin.

One of the most common causes of a false-positive test, especially in a hospital setting, is the recent use of corticosteroids. Steroids, in any form, IV, IM, intrathecal, oral, or inhalational, can suppress the HPA axis and decrease intrinsic cortisol production. Exogenous steroids decrease both baselines as well as adrenal responsiveness to cosyntropin and can cause false-positive results. A detailed history and review of the chart are key in the evaluation of AI and helps in the interpretation of CST. History should include prior glucocorticoid use and use of other drugs that can suppress the HPA axis. It should also include factors affecting cortisol binding, the onset of symptoms, and associated features for reasons noted below.

CST relies on measuring total serum cortisol levels. As with several other hormones, serum cortisol can be found as bound and free. 90% of circulating cortisol is bound to protein. Approximately 70% of it is tightly bound to cortisol binding globulin (CBG), and 10% to 20% is bound loosely to albumin. 10% of cortisol circulates as free and is biologically active. Disorders that reduce (inflammation, rare genetic disorders, Nephrotic syndrome, liver disease, immediate postoperative period, patients requiring intensive care) or increase CBG levels (estrogen, pregnancy, mitotane) need to be considered during interpretation of plasma cortisol. Abnormal levels of CBG and albumin can significantly affect the interpretation of CST and should be taken into account. Unfortunately, no formula or calculation can correct for abnormal CBG or albumin levels, and interpretation relies on clinical judgment.

Free cortisol levels are helpful in patients with abnormal binding globulin. Free cortisol can be measured directly or indirectly in saliva or serum. In critically sick patients, a random free serum cortisol level of more than 1.8 mcg/dL is desired. A stimulated free level of greater than 3.1 mcg/dL is generally accepted during CST.[9] There are several limitations for using free serum cortisol and include decreased availability, the long wait period for results, labor-intensive (need to separate bound and unbound cortisol before measuring free levels). Measuring free cortisol in saliva can be another option in some patients. Salivary cortisol can easily be obtained and correlate well with free cortisol levels. However, data for interpretation in CST is very limited, especially in critically ill patients.[7] There are some methods like Coolen’s method (calculated free cortisol) and free cortisol index (ratio of total cortisol and CBG). However, CBG levels are not readily available and can delay results in limiting practical application.[6]

There are several limitations of low-dose CST. Most studies have utilized high-dose CST and low-dose CST is not validated, especially in critically ill patients. Therefore 1 mcg CST should not be used in the evaluation of AI in critically ill patients.[6] There is also technical difficulty in administering a 1 mcg dose. Cosyntropin vials are supplied as 250 mcg and need dilution by a nurse before administration. Dilution may produce wide variability in results.

Results, Reporting, and Critical Findings

There are no set criteria for diagnosis for AI in critically ill patients, but most agree that in hemodynamically unstable patients with random cortisol levels below 18 mcg/dL, a trial of glucocorticoid therapy is reasonable.[6] Some studies have accepted the rise of cortisol level by greater than 9 mcg/dl from baseline as an acceptable response for 250 mcg cosyntropin in critically ill patients.[4] The major criticism of using delta cortisol levels during CST as criteria especially to diagnose RAI is that most critically ill patients are already maximally stimulated and may not increase cortisol level after administration of cosyntropin.

Studies also show that although CST can be a good indicator of adrenal function in critically ill patients without sepsis, it can be very variable in patients with sepsis. Therefore CST should carefully utilized and interpreted in a critically ill patient. It is of utmost importance that in a critically ill patient suspected of corticosteroid deficiency, the administration of steroids should not be delayed for testing of AI or CST. If clinically feasible, random serum cortisol and ACTH levels should be drawn before administration of stress dose steroids.

Clinical Significance

Diagnosis and treatment of AI in a timely fashion is important in all, but given that it is associated with grave complications, it has extra significance in critically ill patients. It is of utmost importance that in a critically ill patient suspected of corticosteroid deficiency, the administration of steroids should not be delayed for testing of AI or CST. If clinically feasible, random serum cortisol and ACTH should be drawn before the administration of stress dose steroids. 

High-dose CST is also used as a confirmatory test to diagnose non-classic congenital adrenal hyperplasia (NCCAH). NCCAH is characterized by the partial activity of enzyme 21 hydroxylase 21 OH). This leads to the accumulation of 17 hydroxy-progesterone (17OHP) and androgens. The symptoms are much milder than classic congenital hyperplasia (CAH) and are usually limited to menstrual irregularity and symptoms of hyperandrogenism like hirsutism and acne. Cortisol levels are usually normal. This can be diagnosed if baseline 17 OHP is greater than 200 ng/dl. High-dose CST can be used for confirmation if levels are lower than 240nmol/L. During the test,  17 OHP levels are measured at baseline and 60 min after IV/IM administration of 250 mcg cosyntropin. The stimulation of 17OHP to more than 1500 ng/dl confirms the diagnosis of NCCAH.[10]