ACTH-independent macronodular adrenal hyperplasia (AIMAH) is the underlying cause in < 1 % of patients with CS and has also been referred to as massive macronodular adrenocortical disease, autonomous macronodular adrenal hyperplasia, ACTH-independent massive bilateral adrenal disease, and giant macronodular adrenal disease. Age of symptom onset is typically in the fifth or sixth decade, and men have a similar incidence as women [7, 8]. The classic imaging finding is multiple large adrenal nodules with pathology showing two distinct cell types: those containing clear cytoplasm or lipid-rich cells, and those with a compact cytoplasm or lipid-poor cells. It should be noted, however, that the disease may progress at a different rate in each adrenal gland so that what initially appears to be unilateral disease ultimately develops into bilateral nodules. Adrenal cortisol secretion is amplified due to an increased number of adrenocortical cells rather than greater production from individual cells.

Unlike with AIMAH, the actual size of the adrenal gland may be normal but the structure is replaced with multiple nodules scattered throughout an atrophic cortex. On imaging, primary pigmented nodular adrenocortical disease (PPNAD) can have several different appearances ranging from a string of beads to several large nodules. On pathologic examination, most of the nodules are small (< 4 mm) and clearly separate from the surrounding cortex. The nodules seen in PPNAD generally do not show a significant response to exogenous ACTH administration but demonstrate a paradoxical increase in cortisol secretion in response to dexamethasone.

The etiology of PPNAD is evenly split between sporadic and familial causes. The Carney complex is a multiple endocrine neoplasia syndrome involving myxomas (cutaneous and cardiac), spotty skin pigmentation, schwannomas, and several endocrine tumors. PPNAD is the most common endocrine tumor seen in the Carney complex (see Chap. 37 for more information on Carney) and can be found in one fourth of those patients. Additionally, PPNAD has been linked to mutations in the cyclic adenosine monophosphate (cAMP)-dependent protein kinase A ( PRKAR1A) gene [9].

An important limitation of all tests for CS is the difficulty distinguishing true CS from conditions that lead to a temporary increase in cortisol (pseudo-Cushing states). Significant alcoholism, depression, panic attacks, obesity, stress, and intensive care unit (ICU) admission or critical illness can all transiently increase cortisol levels and lead to false-positive results. Additionally, the use of estrogens or oral contraceptives leads to increase in serum cortisol via higher levels of binding proteins (see section on UFC below). When attempting to diagnose CS, the presence of pseudo-Cushing states need to be considered and addressed when possible.

The rationale behind dexamethasone suppression testing is that the normal HPA responds to an increase in corticosteroids by suppressing ACTH secretion which leads to a drop in cortisol. In patients with ACTH-independent CS, the addition of dexamethasone should have minimal impact on serum cortisol because the feedback effect of excess cortisol has already suppressed ACTH secretion. Similarly, ectopic sources of ACTH secretion are generally not inhibited by receiving dexamethasone. Pituitary adenomas also tend not to be completely suppressed by low doses of dexamethasone, though there is often at least some decrease in serum cortisol after administration of dexamethasone. In practice, some patients with adrenal and even ectopic ACTH secretion (EAS) will have cortisol suppression in response to the LDDST. There are several methods for conducting the LDDST but one common approach involves oral administration of 1 mg dexamethasone between 2300 and 2400 h followed by plasma cortisol measurement between 0800 and 0900 h that morning. Overall sensitivity of LDDST may be as high as 95 % with a cutoff point of 1.8 µg/dL, though there is some disagreement over where this threshold should be set. Using a higher cutoff of < 5 µg/dL will decrease sensitivity but fewer false positives results in improved specificity [19–26].

The results of a LDDST can be affected by any of the pseudo-Cushing states mentioned above as well as by any process or medication that alters dexamethasone metabolism. These are discussed in more detail in a later section. Notably, women taking hormone supplementation may have up to a 50 % false-positive rate.

Serum cortisol circulates in an inactive form bound to cortisol-binding protein. The active or free cortisol is a function of both total cortisol levels and concentration relative to binding proteins. Serum measurements are affected by medications and comorbidities that influence cortisol-binding globulin (CBG) levels. These include use of oral contraceptives by women as estrogen increases CBG production and this leads to elevated serum cortisol measurement. By contrast, UFC reflects the quantity of unbound cortisol that is filtered through the kidney and does not fluctuate with CBG concentration. Since UFC is collected over a 24-h period, it provides an overall assessment of free cortisol levels rather than relying on detecting elevated cortisol at a single time point. Sensitivity and specificity vary somewhat depending on the threshold set for an abnormal result but are typically > 85 and 90 %, respectively [27, 28]. There are several limitations to measuring UFC. First, since the assay depends on renal filtration, alterations in kidney function can significantly alter cortisol levels. At a creatinine clearance of < 60 cc/min, the rate of false negatives begins to increase substantially as less cortisol is filtered and collected in the specimen. At the opposite end of the spectrum, a substantial increase in fluid intake can lead to falsely elevated UFC as renal filtration increases. Additionally, performing the test can be something of a burden as patients need to provide a complete 24-h sample for the assay to be reliable.

In otherwise healthy individuals, serum cortisol begins to rise prior to waking and peaks between 7 and 9 a.m. Cortisol then steadily declines throughout the day and reaches a nadir around midnight. This normal circadian rhythm is often interrupted in patients with CS so that midnight cortisol is substantially elevated. Plasma cortisol equilibrates with salivary levels and the concentration is not substantially affected by the amount of saliva produced. Equally important, increases in serum cortisol are rapidly reflected by increases in salivary cortisol. Since measurements of salivary levels do not require needle sticks and blood draws, this assay is a convenient way to measure cortisol in an outpatient setting. When a threshold value of > 145 ng/dL is set, sensitivity and specificity are comparable to the previous assays and range from 92 to 100 % and 93 to 100 %, respectively [27–30].

Salivary glands express 11β-hydroxysteroid dehydrogenase type 2 which converts cortisol to cortisone; it is possible that medications or other substances which block the enzyme may have falsely elevated salivary cortisol. Substances such as licorice and tobacco fall into this category and should be avoided during testing. Since the assay functions based on assumptions of a normal circadian rhythm, patients who work in night shifts or have altered sleep–wake cycles should have the timing of the assay adjusted appropriately or the measurement will miss the actual nadir in cortisol.

If one of the above high sensitivity tests are used to demonstrate elevated cortisol, it is important to confirm the diagnosis with additional testing. Two elevated tests are required to confirm the diagnosis. If the diagnosis is equivocal based on initial testing, additional advanced testing can help better identify patients who truly have CS. One option is use of the 48-h, 2 mg/day LDDST with or without addition of CRH stimulation. This test involves administering 0.5 mg doses of dexamethasone every 6 h for 48 h and then measuring serum cortisol 6 h later. The addition of 1 µg/kg intravenous (IV) ovine CRH 2 h after the last dose of dexamethasone may improve both sensitivity and specificity. The high-dose dexamethasone suppression test (HDDST) is another reasonable alternative though some studies suggest that this test adds little information compared to the LDDST [31]. Alternatively, the midnight serum cortisol test can help confirm the diagnosis of CS [32]. This test requires inpatient admission and can be performed in patients while sleeping or while awake with different thresholds set for each approach.