Endogenous Cushing's syndrome (CS) is a rare disease, and usually characterized by hypertension, diabetes, obesity, osteoporosis, facial rounding, dorsocervical fat pad, thin skin, purple striae, hirsutism, and mood disorders. Efficient diagnostic and screening strategies lead to the diagnosis of a significantly higher number of cases of CS. As a screening test for CS, the Endocrine Society's Clinical Practice Guidelines recommend a single test with a high diagnostic accuracy, among the 1-mg dexamethasone suppression test (1-mg DST), late night salivary cortisol (LNSC), and 24 h urinary free cortisol (UFC). In normal subjects, administering a higher than physiological dose of glucocorticoids prompts the suppression of cortisol secretion. The 1-mg DST explores this normal feedback reaction from the hypothalamic-pituitary-adrenal axis (HPA). It is a simple dynamic test, usually performed in outpatients. A morning serum cortisol level <50 nmol/L suffices to exclude CS, unless there is a strong clinical suspicion to suggest otherwise. The HPA axis reaches a nadir just after a person has fallen asleep, but its circadian rhythm is impaired in CS patients, who feature higher cortisol values at night, which are easy to measure in saliva (the LNSC assay). Saliva collection is also suitable for outpatients since cortisol is stable at room temperature and the collection device can be mailed to the laboratory for analysis. UFC levels reflect the integrated tissue exposure to free cortisol over 24 h, and thus provide a particular picture of endogenous hypercortisolism. In most cases, high UFC levels coincide with severe hypercortisolism. UFC is used not only to diagnose CS, but also to monitor its response to medical treatment. All screening tests have procedural snares: some drugs can interfere with the DST; false-positive or false-negative LNSC results may be due to an inadequate soaking of the device or to cyclic CS; and in the case of UFC it is important to ensure that patients provide complete urine collections with appropriate total volumes. Measuring cortisol with antibody-based immunoassays can also generate false-positive results due to cross-reactivity between cortisol, cortisone and other metabolites. Structurally-based assays, such as liquid chromatography with tandem mass spectrometry, only measure cortisol and have only recently become available for use in routine clinical practice. This review summarizes the recent literature on the clinical and biochemical aspects of CS diagnostics with a view to helping physicians choose the best screening test for diagnosing endogenous hypercortisolism.
