Laboratory assessment of hypothalamic-pituitary-adrenal axis function

INTRODUCTION — 

The hypothalamic-pituitary-adrenal (HPA) axis regulates cortisol production. Hypothalamic secretion of corticotropin-releasing hormone (CRH) and vasopressin into the hypothalamic-pituitary portal veins stimulates the release of corticotropin (ACTH) from the pituitary gland. ACTH, in turn, promotes cortisol secretion by the adrenal cortex. Measurements of cortisol and its secretagogue ACTH are critical in the diagnostic and etiologic evaluation of hypercortisolism and adrenal insufficiency. However, laboratory measurements are influenced by many factors that can complicate the interpretation of results, including the episodic and diurnal variation in hormone secretion and the different assays used for hormone measurement.

This topic will review the laboratory assessment of HPA axis function, clinical indications for these measurements, and potential pitfalls and caveats for the interpretation of test results. Detailed approaches to the diagnosis of Cushing syndrome and adrenal insufficiency are discussed separately. (See "Establishing the diagnosis of Cushing syndrome" and "Determining the etiology of adrenal insufficiency in adults".)

Laboratory assessment of adrenal aldosterone production, which is not regulated by the HPA axis, is also discussed separately. (See "Diagnosis of primary aldosteronism" and "Assays of the renin-angiotensin-aldosterone system".)

BASIC PHYSIOLOGY OF THE HYPOTHALAMIC-PITUITARY-ADRENAL AXIS — 

The adrenal cortex consists of three functional zones, two of which are regulated by the hypothalamic-pituitary-adrenal (HPA) axis [1,2]:

●The zona fasciculata (middle layer of the adrenal cortex) secretes cortisol and is regulated by the corticotropin-releasing hormone (CRH)-corticotropin (ACTH) system.

●The zona reticularis (innermost layer of the adrenal cortex) secretes dehydroepiandrosterone (DHEA), DHEA sulfate (DHEAS), and smaller amounts of androstenedione, testosterone, and other 19-carbon steroids and is primarily regulated by the CRH-ACTH system. Proopiomelanocortin (POMC)-derived peptides other than ACTH also may contribute to the regulation of adrenal androgen production.

●The zona glomerulosa (outermost layer of the adrenal cortex) secretes aldosterone and is primarily regulated by the renin-angiotensin system and plasma potassium.

In normal physiology, secretion of both ACTH and cortisol exhibits a pulsatile, diurnal rhythm (figure 1). Peak levels occur a few hours before awakening, followed by pulsatile, decreasing values throughout the day and nadir values in the evening and early part of the sleep period [3]. Although cortisol secretion generally reflects ACTH secretion, cortisol secretion lags by a few minutes, and cortisol has a longer disappearance half-life (approximately 80 versus eight minutes for ACTH). Consequently, cortisol excursions are dampened relative to those of ACTH. This diurnal rhythm informs clinical testing strategies, and time-specific reference ranges are critical for interpreting cortisol and ACTH measurements. (See 'Serum cortisol' below and 'Salivary cortisol' below and 'ACTH' below.)

POMC is the ACTH precursor peptide. It is also the source of beta-lipotropin (beta-LPH). ACTH and beta-LPH are cleaved further to yield alpha- and beta-melanocyte-stimulating hormone (alpha-MSH and beta-MSH), beta-endorphin, and met-enkephalin. Neither POMC nor these other POMC-derived hormones have clinical indications for measurement. Similarly, plasma CRH is not measured clinically. (See 'CRH' below.)

CORTISOL AND CORTISONE

Serum cortisol — Serum cortisol assays measure total cortisol, which includes both the free and protein-bound fractions. Approximately 90 percent of circulating cortisol is protein bound. Although methods have been developed for measuring serum free cortisol [4-8], these are not widely used. Measurement or calculation of free cortisol levels has been used in research settings, particularly in the context of sepsis and critical illness [9-11]. Salivary cortisol can be used as a surrogate for serum free cortisol [12,13]. (See 'Salivary cortisol' below and "Diagnosis of adrenal insufficiency in adults", section on 'Critical illness'.)

Clinical use — A morning serum cortisol value may help to exclude adrenal insufficiency, but it is not a sensitive test for either cortisol deficiency or excess. Cortisol secretion is episodic, and morning reference intervals for serum cortisol are typically broad and assay specific. A single serum cortisol value that is within the reference interval is inconclusive, and additional testing is usually required to confirm a diagnosis.

●Adrenal insufficiency – Patients with primary or central (ie, secondary or tertiary) adrenal insufficiency occasionally have extremely low early morning serum cortisol concentrations. Serum cortisol values obtained at other times of day have no role in the diagnostic evaluation of chronic adrenal insufficiency, which is reviewed in detail separately. (See "Diagnosis of adrenal insufficiency in adults", section on 'Suspected chronic adrenal insufficiency'.)

●Cushing syndrome – In the evaluation of Cushing syndrome, bedtime serum cortisol measurement can be helpful but is rarely used due to inconvenience [14]. Morning serum cortisol concentrations are typically within or slightly above the reference range and therefore are not useful for diagnosis. The diagnostic evaluation of Cushing syndrome is reviewed in detail separately. (See "Establishing the diagnosis of Cushing syndrome", section on 'Available tests'.)

Assay methods — Various assay methods are used to measure serum cortisol [3] and vary in specificity. Among available assays, structurally based assays (liquid chromatography-tandem mass spectrometry [LC-MS/MS]) are the most specific for the cortisol molecule [15]. Assay methods also include monoclonal and polyclonal antibody-based immunoassays. Due to the variable specificity across these assay types, assay-dependent diagnostic thresholds are used for serum cortisol levels in the evaluation of adrenal insufficiency. (See "Diagnosis of adrenal insufficiency in adults", section on 'Suspected chronic adrenal insufficiency'.)

●Immunoassays – Platform automated analyzers commonly use fluorescent, chemiluminescent, and other labels in place of radioisotopic labels. Newer generation immunoassays using monoclonal antibodies give results more similar to LC-MS/MS than previous generations of immunoassay [16].

●Structurally based assays – The development of high-throughput techniques to simultaneously measure multiple samples makes these labor-intense assays feasible for commercial use [17]. Structurally based assay methods separate cortisol from other steroids and steroid metabolites; cortisol is then measured by mass spectrometry [18].

Caveats for interpretation — Certain medications, nutritional status, comorbid conditions, and age all can affect serum cortisol levels and therefore should be considered in the interpretation of serum cortisol values.

●Recent exposure to exogenous glucocorticoids – Exogenously administered glucocorticoids can alter serum cortisol values through various mechanisms. If an antibody-based assay is used, glucocorticoids that cross-react with the assay antibody can lead to spurious cortisol elevations. Hydrocortisone is molecularly identical to cortisol and therefore is detected by all cortisol assays. Whether cortisone is detected depends on the specific assay method.

Exogenous glucocorticoids also can suppress the hypothalamic-pituitary-adrenal (HPA) axis, leading to central adrenal insufficiency and low cortisol values [19]. Although this suppression is typically evident only when glucocorticoid doses exceed prednisone 5 mg daily or its equivalent, HPA axis sensitivity to exogenous glucocorticoid exposure is highly variable across individuals; as a result, lower oral glucocorticoid doses or glucocorticoids administered through other routes (eg, inhaled, topical) may cause HPA axis suppression. (See "Determining the etiology of adrenal insufficiency in adults", section on 'History of exogenous glucocorticoid use'.)

●Nutritional status and body weight – Severe underweight, fasting, eating disorders, and obesity can lead to physiologic (non-neoplastic) hypercortisolism with elevated mean serum cortisol levels. In the absence of underweight, the effects of caloric restriction on serum cortisol have been less consistent, and increases in serum cortisol may be limited to the first few weeks of caloric restriction [20].

●Abnormal corticosteroid-binding globulin (CBG) – Diagnostic thresholds for serum cortisol levels assume a normal concentration of CBG. Low CBG concentrations, as may be seen in genetic deficiency, cirrhosis, critical illness, or nephrotic syndrome, result in lower total serum cortisol values [10,21]. Conversely, high CBG concentrations increase serum cortisol concentrations [21]. In such cases, 24-hour urinary free cortisol and salivary cortisol better reflect HPA axis function [22-24]. Causes of altered CBG levels and the diagnostic evaluations for adrenal insufficiency and Cushing syndrome in individuals with abnormal CBG levels are reviewed separately. (See "Diagnosis of adrenal insufficiency in adults", section on 'Abnormal CBG' and "Establishing the diagnosis of Cushing syndrome", section on 'Available tests'.)

●Depression – Major depressive disorders, particularly severe depression with melancholic features, can result in physiologic (non-neoplastic) hypercortisolism that may be difficult to differentiate from neoplastic hypercortisolism (Cushing syndrome) [25-28]. (See "Establishing the diagnosis of Cushing syndrome", section on 'Bedtime serum cortisol'.)

●Non-glucocorticoid drugs – Several drugs induce hepatic cytochrome P450 enzymes that metabolize steroids (table 1). Barbiturates, phenytoin, rifampin, aminoglutethimide, and mitotane increase the metabolic clearance of steroids. They have a preferential effect on synthetic 9-fluoro steroids (eg, dexamethasone and fludrocortisone) as compared with natural steroids. These drugs generally do not alter basal serum cortisol concentrations; an exception is mitotane, which also increases CBG and therefore raises total cortisol. However, these drugs can interfere with dexamethasone suppression and metyrapone stimulation tests. Such medications also may necessitate an increased steroid replacement dose in patients with adrenal insufficiency. (See "Dexamethasone suppression tests", section on 'Sources of error' and "Treatment of adrenal insufficiency in adults", section on 'Special considerations for dosing'.)

●Chronic alcohol use – Chronic, excessive alcohol use can cause non-neoplastic hypercortisolism (formerly known as pseudo-Cushing syndrome) and high serum cortisol concentrations [28,29]. Occult or underreported alcohol use can be detected by measuring blood phosphatidylethanol (PEth) available as a routine clinical laboratory test [29-31]. As with other causes of physiologic hypercortisolism, a circadian pattern of cortisol secretion is generally preserved, with a bedtime nadir in serum cortisol. (See "Establishing the diagnosis of Cushing syndrome", section on 'Bedtime serum cortisol'.)

●Liver and kidney dysfunction – Severe liver dysfunction (ie, cirrhosis) can lead to reduced serum cortisol concentrations in part through loss of hepatic synthetic function and lower CBG levels. Further, adrenal insufficiency during critical illness is common in individuals with cirrhosis, a phenomenon sometimes called "hepatoadrenal syndrome" [32,33]. The diagnosis of hepatoadrenal syndrome is challenging, as conventional diagnostic thresholds for basal and stimulated cortisol levels cannot be used in the setting of low CBG. Thus, hepatoadrenal syndrome remains primarily a clinical diagnosis. In such individuals, whether glucocorticoid treatment during critical illness improves outcomes remains uncertain [32,33]. The evaluation for adrenal insufficiency in individuals with abnormal CBG is reviewed separately. (See "Diagnosis of adrenal insufficiency in adults", section on 'Abnormal CBG'.)

Chronic kidney disease (CKD) and end-stage kidney disease (ESKD) can cause ACTH-driven non-neoplastic hypercortisolism [28,30,34]. This physiology may be detected by an increased bedtime salivary cortisol level [34-36]. Dexamethasone metabolism is altered in CKD, so the low-dose dexamethasone suppression test may generate misleading results and requires the simultaneous measurement of serum dexamethasone concentrations [37]. (See 'Salivary cortisol' below and "Establishing the diagnosis of Cushing syndrome", section on 'Other'.)

●Thyroid dysfunction – Thyroid hormone regulates the rate of cortisol metabolism. If HPA feedback mechanisms are intact, serum cortisol concentrations should remain within normal limits in hyperthyroidism, as increased cortisol production compensates for accelerated clearance [38]. In patients with adrenal insufficiency, however, either iatrogenic or endogenous hyperthyroidism can precipitate adrenal crisis. In individuals with hyperthyroidism due to Graves' disease, acute stress may lead to increased serum levels of both cortisol and thyroid hormone. (See "Pathogenesis of Graves' disease", section on 'Stress'.)

In primary hypothyroidism, reduced cortisol clearance may raise mean serum cortisol levels and therefore potentially mask mild adrenal insufficiency [39,40].

●Age – Term neonates may acquire a relatively normal and stable cortisol circadian rhythm early in postnatal life and even by one month of age [41]. Whereas serum cortisol levels remain relatively stable over young adulthood and midlife, an increase in late-night salivary cortisol and mean daily serum cortisol may be evident during older age [42,43].

Salivary cortisol — Salivary cortisol has become one of the most commonly used approaches to the diagnosis of hypercortisolism [14]. Serum free cortisol diffuses freely into saliva. Therefore, measurements of salivary cortisol more accurately reflect serum free (biologically active) cortisol concentrations than do measurements of serum total cortisol [13]. The salivary cortisol concentration is independent of salivary flow rate [44,45].

Clinical use — Salivary cortisol (and cortisone, if available) is used most often in the diagnostic evaluation for Cushing syndrome [12,13,46,47] and, more recently, in the evaluation for adrenal insufficiency.

●Cushing syndrome – In Cushing syndrome, bedtime salivary cortisol concentrations are elevated. Measuring salivary cortisol is useful for serial measurements in ambulatory patients, who can collect and store multiple samples and then return all collected samples to a clinical laboratory. Serial salivary cortisol measurement therefore is particularly helpful in the evaluation of patients with suspected cyclical Cushing syndrome [48-52]. Salivary cortisol can detect postoperative recurrence of Cushing disease, which may occur even 10 years after surgical remission [53,54].

Concurrent measurement of salivary cortisone may improve the performance of the test for suspected Cushing syndrome and also helps to identify sample contamination with topical hydrocortisone [55]. A bedtime salivary cortisol level below the upper limit of the reference range is useful to exclude Cushing syndrome [46]. However, an elevated level is not sufficient to establish the existence of neoplastic hypercortisolism and requires further evaluation, ideally performed under specialist care [28,30,34]. (See "Establishing the diagnosis of Cushing syndrome", section on 'Bedtime salivary cortisol'.)

●Adrenal insufficiency – In adrenal insufficiency, morning salivary cortisol concentrations are decreased. Although not commonly used for the diagnosis of adrenal insufficiency, morning salivary cortisol may be used for diagnosis, particularly in individuals with abnormal CBG levels [56,57]. Salivary cortisol sampled throughout the day also has been used as a strategy to monitor glucocorticoid replacement therapy [58,59]. (See "Diagnosis of adrenal insufficiency in adults", section on 'Tests not affected by abnormal CBG'.)

Assay methods — The most common approaches to measure salivary cortisol are immunoassay and LC-MS/MS [13,15,60,61]. The reference ranges and units reported for the many different methods have not been harmonized; therefore, clinicians should use the reference range accompanying results obtained from their clinical laboratory.

Saliva can be obtained using a cotton tube or polyester swab or by collecting saliva into a plain tube. Collected samples can be stored at room temperature for many days [62] or frozen for extended periods. LC-MS/MS is more specific and can also measure salivary cortisone.

Urinary free cortisol — Measuring daily urinary free cortisol (UFC) excretion provides an integrated index of serum free cortisol concentration over a period of 24 hours, whereas measurements of serum or salivary cortisol only provide information about an instant in time.

Clinical use — Measurement of 24-hour UFC is primarily used in the diagnosis of Cushing syndrome. UFC measures only non-protein-bound cortisol, which is filtered by the kidney unchanged. Therefore, UFC is unaffected by medications and medical conditions that alter CBG or albumin. In healthy individuals without excess fluid intake, UFC ranges from approximately 10 to 55 mcg/day (27 to 150 nmol/day), as measured by specific immunoassays or LC-MS/MS. (See 'Assay methods' below and 'Caveats for interpretation' below.)

●Cushing syndrome – While UFC is recommended as a screening test for Cushing syndrome, it is a less sensitive test than the 1 mg dexamethasone suppression test for detection of mild hypercortisolism due to mild autonomous cortisol secretion (MACS) [63]. UFC is also less sensitive than bedtime salivary cortisol for detecting recurrence of Cushing disease [54].

Serum total cortisol concentrations exceed the binding capacity of CBG at approximately 25 mcg/dL (690 nmol/L), reflecting the normal circadian peak level; above this concentration serum free cortisol concentrations increase rapidly, leading to a marked increase in urinary cortisol excretion. Use of the 24-hour UFC test in the diagnostic evaluation of Cushing syndrome is discussed in detail separately. (See "Establishing the diagnosis of Cushing syndrome", section on '24-hour urinary cortisol excretion'.)

●Monitoring medical therapy for Cushing syndrome – UFC is also used to monitor the efficacy of steroidogenesis inhibitors for the medical management of Cushing syndrome. Agents that inhibit CYP11B1 (metyrapone and osilodrostat) increase circulating and urinary levels of 11-deoxycortisol, so LC-MS/MS assays should be used to avoid immunoassay cross-reactivity with 11-deoxycortisol [64]. (See "Medical therapy of hypercortisolism (Cushing syndrome)", section on 'Adrenal steroidogenesis inhibitors'.)

When to avoid use

●Suspected adrenal insufficiency – Urinary cortisol excretion is low in patients with primary and central adrenal insufficiency [65]. However, urinary cortisol values in patients with adrenal insufficiency overlap with the lower part of the normal reference range. Therefore, UFC alone is not reliable for diagnosis.

●Titrating glucocorticoid replacement therapy – Urinary cortisol measurements are not a good metric for titrating cortisol replacement therapy [66]. A key limitation of this strategy is that the number of daily doses of hydrocortisone will affect UFC values. Thus, the same daily replacement dose of hydrocortisone (eg, 30 mg) results in a greater increase in 24-hour UFC excretion when given as a single dose than when given in multiple divided doses. A single dose of hydrocortisone is absorbed within 20 to 30 minutes. As a result, serum cortisol concentrations may temporarily exceed the binding capacity of CBG, and serum free cortisol concentrations transiently rise. This free cortisol is excreted in the urine until the serum total cortisol concentration, falling rapidly, reaches the binding capacity of CBG. When multiple doses are used to provide the same total amount of hydrocortisone, serum cortisol concentrations may never exceed the binding capacity of CBG. Consequently, the 24-hour UFC excretion is lower and better represents the integrated serum free cortisol concentrations. Titrating glucocorticoid replacement therapy is reviewed in detail separately. (See "Treatment of adrenal insufficiency in adults", section on 'Dose titration'.)

Assay methods — Various assay methods are used to measure UFC, and reference ranges vary both across and within the different analytic methods. Most reference laboratories use LC-MS/MS. UFC measurement through structurally based assays (eg, LC-MS/MS) do not use antibodies. These methods are more specific than immunoassays, have a lower reference range (10 to 55 mcg/day [27 to 150 nmol/day]), and correlate well with each other [67].

Antibody-based immunoassays are relatively nonspecific and may cross-react with other steroids and steroid metabolites (eg, prednisolone, 6-beta-hydroxycortisol) [67-73]. This lower specificity is typically reflected by a higher upper limit of the reference range. Nonspecific urine cortisol immunoassays have reference ranges of 20 to 90 to 100 mcg/day (55 to 250 to 285 nmol/day) or greater, considerably higher than the reference ranges for more specific immunoassays and LC-MS/MS.

Caveats for interpretation

●Need for proper collection – The validity of UFC measurement requires both an appropriately timed urine collection and avoidance of excessive fluid intake. Over or under-collection of the 24-hour specimen can bias UFC results. Concurrent measurement of creatinine excretion is essential and helps verify proper collection based on normal, age- and sex-adjusted reference ranges for creatinine excretion. For example, in adults under the age of 50 years, daily creatinine excretion should be 20 to 25 mg (177 to 221 micromol) per kg lean body weight in males and 15 to 20 mg (133 to 177 micromol) per kg lean body weight in females. From the ages of 50 to 90 years, creatinine excretion progressively declines due largely to decreasing muscle mass. (See "Calculation of the creatinine clearance".)

Importantly, creatinine excretion cannot be used to correct UFC in case of improper collection. The excretion of cortisol and its metabolites varies diurnally, whereas creatinine excretion does not, so both over- and under-collection will change the ratio between creatinine and cortisol excretion. Excessive fluid intake during urine collection similarly can affect results, as cortisol but not creatinine excretion increases with higher fluid consumption [74,75]. The importance of appropriately timed collection and avoidance of high fluid intake (eg, ≥3 to 4 liters daily) should be stressed to the patient. (See "Patient education: Collection of a 24-hour urine specimen (Beyond the Basics)".)

●Topical hydrocortisone use – Another potential problem with UFC is contamination of the urine sample with hydrocortisone cream or ointment applied to the vulva [76,77]. As with salivary cortisol, contamination of the sample with hydrocortisone [55] will falsely increase UFC regardless of the assay method.

●Impaired kidney function – The validity of UFC measurement also requires normal kidney function [78]. Cortisol excretion falls progressively over moderate (creatinine clearance <60 mL/min) to severe (creatinine clearance <20 mL/min) kidney impairment [78]. Therefore, in individuals with moderate to severe CKD and suspected Cushing syndrome, alternative diagnostic tests should be used. (See "Establishing the diagnosis of Cushing syndrome", section on 'Available tests'.)

●Recent/concomitant stress – Urinary cortisol excretion is also disproportionately high after major episodes of endogenous cortisol secretion, as may occur during acute or chronic stress. Consequently, in the diagnostic evaluation of Cushing syndrome, UFC should be measured outside of episodes of major physiologic stress.

Urinary cortisone — Cortisone can be measured in urine by LC-MS/MS [79-81]. Its daily excretion is increased in patients with Cushing syndrome due to endogenous cortisol production. Measurement of urinary cortisone may be combined with urinary cortisol to aid in the diagnosis of factitious Cushing syndrome due to ingestion of hydrocortisone. In such patients, urinary cortisol excretion is high, but urinary cortisone excretion is low.

Urinary cortisone also can used in the diagnosis of the syndrome of apparent mineralocorticoid excess. In individuals with clinical evidence of this syndrome, an elevated ratio of UFC to free urinary cortisone supports the diagnosis [82,83]. (See "Apparent mineralocorticoid excess syndromes (including chronic licorice ingestion)", section on 'Syndrome of apparent mineralocorticoid excess'.)

CORTICOTROPIN (ACTH) AND CORTICOTROPIN-RELEASING HORMONE (CRH)

ACTH

Clinical use — The plasma ACTH level is used clinically to discriminate between primary and central adrenal insufficiency and between ACTH-dependent and -independent forms of Cushing syndrome. The time of day at which the sample is collected is an important determinant of the utility of plasma ACTH measurement, which is often best interpreted with simultaneous measurement of serum cortisol.

●Adrenal insufficiency – In individuals with established cortisol deficiency, a morning plasma ACTH level helps distinguish between primary and central adrenal insufficiency. Plasma ACTH should be measured in the morning, if possible, when plasma ACTH and serum cortisol concentrations are usually at their highest [65]. When clinical suspicion for adrenal insufficiency is high, serum cortisol and plasma ACTH levels may be measured simultaneously as an initial diagnostic test. The interpretation of ACTH values in the etiologic evaluation of adrenal insufficiency is reviewed separately. (See "Determining the etiology of adrenal insufficiency in adults", section on 'Establish the level of defect'.)

●Cushing syndrome – In the etiologic evaluation of Cushing syndrome, ACTH can be measured at any time of day [84]. (See "Establishing the diagnosis of Cushing syndrome".)

•ACTH independent – In patients with Cushing syndrome due to adrenocortical tumors, bilateral micronodular dysplasia, or other primary adrenocortical disorders (ie, ACTH-independent Cushing syndrome), plasma ACTH concentrations are usually suppressed and can be undetectable. Nonetheless, if hypercortisolism is mild or cyclical, plasma ACTH may not be overtly suppressed. (See "Establishing the cause of Cushing syndrome", section on 'Is hypercortisolism ACTH-dependent or independent?' and "Establishing the cause of Cushing syndrome".)

•ACTH dependent – In patients with Cushing disease or ectopic ACTH syndrome (ie, ACTH-dependent Cushing syndrome), bedtime plasma ACTH and serum cortisol concentrations are both high. The values for both tend to be higher in ectopic ACTH syndrome than in Cushing disease, but considerable overlap exists. The morning plasma ACTH levels are often within the reference range but nonetheless inappropriately high for the degree of hypercortisolism. (See "Establishing the cause of Cushing syndrome", section on 'ACTH-dependent CS (ACTH >20 pg/mL)' and "Establishing the cause of Cushing syndrome".)

Assay method — Commercially available two-site immunometric assays for plasma ACTH have an analytical sensitivity between 0.6 and 9 pg/mL (0.12 to 2.0 pmol/L) [85]. However, performance differs among assays and should be considered when interpreting patient results [86-88].

The importance of assay method is illustrated by erroneously elevated plasma ACTH values found with the Siemens ACTH Immulite assay, in some cases leading to inappropriate diagnostic procedures and treatment [89-91]. Accordingly, the possibility of measurement error should always be considered, particularly when laboratory data are surprising or inconsistent. Many reference laboratories have switched to the Roche Cobas platform as a result. (See 'Potential errors in assay measurements' below.)

Caveats for interpretation — Several factors may influence plasma ACTH assay results, including the episodic nature of ACTH secretion, clinical context, and the need for proper sample collection and storage.

●Episodic secretion – ACTH secretion in primary adrenal insufficiency, congenital adrenal hyperplasia, and ACTH-dependent Cushing syndrome is episodic. This episodic secretion influences ACTH measurement more than that of cortisol because of the shorter plasma disappearance half-time for ACTH (ie, approximately eight versus 80 minutes). ACTH measurement in more than two samples therefore may be useful, particularly if the initial result is equivocal. Serial sampling can be performed at the same time on multiple days or by obtaining multiple samples 30 to 60 minutes apart on the same day.

●Sample collection and storage – ACTH may be unstable in blood at room temperature. It is cleaved by enzymes in blood cells and platelets and adheres to glass and some plastic surfaces [89]. Therefore, sample collection, processing, and storage may affect the measured ACTH concentration; the severity of this problem varies with the antibodies used in the assays [92]. In general, ACTH blood samples should be drawn into plastic tubes with ethylenediaminetetraacetic acid (EDTA) and kept on ice until centrifugation, separation, and freezing of the plasma sample.

●Recent stress – Both healthy individuals and those with adrenal disorders may respond rapidly to stress with increased ACTH secretion. For this reason, in the evaluation of Cushing syndrome, blood samples should be obtained during an unstressed state. In contrast, major physiologic stress may be helpful as a provocative test in the evaluation of adrenal insufficiency. (See "Diagnosis of adrenal insufficiency in adults", section on 'Suspected adrenal crisis'.)

●Depression and other forms of non-neoplastic hypercortisolism – Patients with major depressive disorder, particularly major depression with melancholic features, may have increased plasma ACTH concentrations [25]. Non-neoplastic hypercortisolism also has been found in patients with obstructive sleep apnea, chronic kidney disease (CKD), and alcohol use disorder [30]. Thus, physiologic hypercortisolism should be excluded prior to ACTH measurement in individuals with depression. (See "Establishing the diagnosis of Cushing syndrome", section on 'Exclude physiologic hypercortisolism'.)

●Exogenous glucocorticoid use – Current or recent glucocorticoid use must be excluded prior to ACTH measurement, as exogenous glucocorticoids can acutely or chronically suppress hypothalamic-pituitary-adrenal (HPA) function. The likelihood of ACTH suppression depends on the dose and duration of glucocorticoid use, but individuals vary in HPA axis sensitivity to exogenous glucocorticoid exposure. (See "Determining the etiology of adrenal insufficiency in adults", section on 'History of exogenous glucocorticoid use'.)

●Detection of unprocessed ACTH or ACTH fragments – Two-site "sandwich" immunoassays for ACTH use two different monoclonal or affinity-purified polyclonal antibodies. These antibodies may not react with proopiomelanocortin (POMC) or forms intermediate between POMC and ACTH [85,92,93]. This specificity may be a disadvantage for diagnosing ectopic ACTH syndrome, in which a large percentage of circulating immunoreactive ACTH may represent incompletely processed or unprocessed POMC. No antibody for POMC is commercially available.

In addition, excessive concentrations of ACTH fragments, such as ACTH (1-24), can compete for binding to one or the other of the two antibodies, preventing intact ACTH from coupling to both antibodies and therefore causing factitiously low ACTH concentrations [94]. This is unlikely to be a problem in measuring endogenous ACTH. However, this phenomenon can cause confusion when a blood sample is drawn (typically in error) after injection of a pharmacologic dose of cosyntropin (synthetic ACTH (1-24) [94].

CRH — Measurement of serum corticotropin-releasing hormone (CRH) is not currently available in reference laboratories.

POTENTIAL ERRORS IN ASSAY MEASUREMENTS

●When to suspect measurement error – The possibility of preanalytical error or laboratory measurement error should be considered whenever reported laboratory values are inconsistent or highly unexpected based on clinical presentation. Tests that produce conflicting or unexpected results always should be repeated, and diagnosis should be deferred until inconsistent findings are reconciled. In case of inconsistent or unexpected results, patients also should be queried about use of any medications or supplements that could interfere with laboratory assays, including any inhaled, intranasal, topical, or injectable glucocorticoid-containing medications.

●Laboratory assay error or interference – With the development of newer generations of cortisol immunoassays using monoclonal antibodies [16,95,96], the performance of platform immunoassays has dramatically improved, and most have a good correlation with liquid chromatography-tandem mass spectrometry (LC-MS/MS) results [97,98]. Furthermore, as LC-MS/MS methods become more standardized, their performance is much more reliable and reproducible. Contamination with exogenous cortisol (hydrocortisone) cannot be detected in any assay method but can be identified with the simultaneous measurement of a nonelevated concentration of salivary cortisone [55,99]. (See "Diagnosis of adrenal insufficiency in adults", section on 'Influence of cortisol assay technique'.)

●Other potential sources of error – In addition to laboratory assay error, inaccurate test results may be due to improper sample collection or processing (preanalytical error). Errors in sample collection and/or processing are most likely for measurements of 24-hour urinary free cortisol (UFC) and of plasma corticotropin (ACTH). (See 'Caveats for interpretation' above and 'Caveats for interpretation' above.)

SOCIETY GUIDELINE LINKS — 

Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Diagnosis and treatment of Cushing syndrome" and "Society guideline links: Adrenal insufficiency".)

SUMMARY

●Assessment of hypothalamic-pituitary-adrenal (HPA) axis function – Measurements of cortisol and its secretagogue corticotropin (ACTH) are extremely useful in the diagnostic and etiologic evaluation of hypercortisolism and adrenal insufficiency. However, laboratory measurements are influenced by many factors that can complicate the interpretation of results, including the episodic and diurnal variations in hormone secretion and the different assays used for measurement. (See 'Introduction' above and 'Basic physiology of the hypothalamic-pituitary-adrenal axis' above.)

●Serum cortisol – Morning serum cortisol values can be helpful as initial screening tests for adrenal insufficiency. Nonetheless, cortisol secretion is episodic, and morning reference ranges for serum cortisol are typically broad and assay specific. A single serum value that falls within the reference range is inconclusive, and additional testing is usually required to confirm a diagnosis. (See 'Clinical use' above and 'Caveats for interpretation' above and "Diagnosis of adrenal insufficiency in adults", section on 'Basal serum cortisol testing'.)

●Salivary cortisol – Measurements of salivary cortisol more accurately reflect serum free (biologically active) cortisol concentrations than do measurements of serum total cortisol. In Cushing syndrome, bedtime salivary cortisol concentrations are elevated. Salivary cortisol is valuable for both initial diagnosis and monitoring patients for postsurgical recurrence of neoplastic hypercortisolism. (See "Establishing the diagnosis of Cushing syndrome", section on 'Bedtime salivary cortisol'.)

In adrenal insufficiency, morning salivary cortisol concentrations are decreased and may be particularly useful for diagnosis in individuals with abnormal levels of corticosteroid-binding globulin (CBG). (See 'Clinical use' above and "Diagnosis of adrenal insufficiency in adults", section on 'Tests not affected by abnormal CBG'.)

●Urinary free cortisol (UFC) – Measurement of 24-hour UFC is primarily used in the diagnosis of Cushing syndrome. Disadvantages include the need for 24-hour urine collection and lower sensitivity compared with late-night salivary cortisol or the low-dose dexamethasone suppression test. UFC measures only free cortisol and is unaffected by medications and medical conditions that affect CBG. UFC is also used to monitor the efficacy of steroidogenesis inhibitors used as medical therapy for Cushing syndrome. (See 'Clinical use' above and "Establishing the diagnosis of Cushing syndrome", section on 'Available tests'.)

UFC should not be used to diagnose adrenal insufficiency or to titrate glucocorticoid replacement therapy in individuals with known adrenal insufficiency. (See 'When to avoid use' above.)

●Plasma ACTH – The plasma ACTH level is used clinically to discriminate between primary and central (ie, secondary or tertiary) adrenal insufficiency and between ACTH-dependent and -independent forms of Cushing syndrome. (See 'Clinical use' above and "Determining the etiology of adrenal insufficiency in adults", section on 'Establish the level of defect' and "Establishing the cause of Cushing syndrome", section on 'Measure plasma ACTH'.)

●Potential errors in measurement – The possibility of measurement error should be considered whenever reported laboratory values are inconsistent or highly unexpected based on clinical presentation. (See 'Potential errors in assay measurements' above and "Diagnosis of adrenal insufficiency in adults", section on 'Influence of cortisol assay technique'.)

ACKNOWLEDGMENT — 

The views expressed in this topic are those of the author(s) and do not reflect the official views or policy of the United States Government or its components.