Excluding Cushing’s syndrome: saliva, urine or serum cortisol?

Cushing’s syndrome may be suspected in patients with classical features (fat redistribution, proximal myopathy, easy bruising, facial plethora, red/purple striae), hypertension (particularly where a secondary cause is under consideration) or accelerated osteoporosis or diabetes mellitus. After iatrogenic Cushing’s syndrome due to exogenous glucocorticoid use is excluded, biochemical testing is performed.

Dr Michael Page, Chemical Pathologist, Jandakot

A key early biochemical feature of endogenous Cushing’s syndrome is loss of the normal diurnal variation of cortisol secretion. Usually, cortisol secretion is highest in the morning and lower in the evening. In patients with Cushing’s syndrome, this rise and fall is flattened, leading to higher than expected concentrations in the evening, but morning concentrations are frequently unaffected. 

Hence, the measurement of serum cortisol in the morning (other than in the context of an overnight dexamethasone suppression test; see later) is highly insensitive and of little value in most cases. The main utility of morning serum cortisol measurement is in ruling out adrenal insufficiency.

The three tests that are useful in the initial evaluation of suspected Cushing’s syndrome are late night salivary cortisol (LNSC), 24-hour urine free cortisol (UFC), and the overnight dexamethasone suppression test (ONDST). Each derives its sensitivity from the biochemical characteristics of Cushing’s syndrome. 

If suspicion is high, two different tests should be performed. If it is low, two negative LNSC results, two negative UFC results, or one negative ONDST result is generally sufficient to rule out the diagnosis. 

While some doctors may be unfamiliar with LNSC, it is an excellent first choice. It is more convenient for the patient than a 24-hour urine collection, less prone to protocol errors or medication interference than the ONDST and has the required performance characteristics to rule out Cushing’s syndrome in most patients. 

The doctor simply provides the patient with two forms requesting “late night salivary cortisol”. The collection centre provides a special collection device to the patient, along with written instructions. At 11pm, the patient removes a small swab from the device, holds it in their mouth for two minutes then places it back into the device. If done on two consecutive evenings, the patient can return both collection devices to the collection centre during business hours, provided that they are kept in a refrigerator at home in the interim. 

Similarly, if UFC is chosen as an initial test, the collection canisters and instructions are provided by the collection centre. As this test measures the total free cortisol excreted in urine across a 24-hour period, it is less susceptible to altered sleep-wake cycles than LNSC. Therefore, it may be a better choice for shift-workers. 

Figure: Rationale for preferred tests for Cushing’s syndrome (Original figure by MMP)

In the case of the ONDST, the patient takes 1mg dexamethasone orally at 11pm, and returns for serum cortisol measurement the following morning at 8am. 

Some laboratories provide the dexamethasone and instructions; others require the patient to obtain the dexamethasone on a prescription. 

This test is preferable for excluding autonomous cortisol secretion in patients with adrenal incidentalomas. However, false-positive results can be caused by oestrogens (including the combined oral contraceptive pill), medications that accelerate the metabolism of dexamethasone, such as some anticonvulsants, and mistiming of the dexamethasone dose in relation to the blood test.

Any abnormal result would usually prompt referral to an endocrinologist. Further advice on testing can also be provided by a chemical pathologist at your preferred laboratory.

Key messages
  • Careful selection of initial biochemical tests for Cushing’s syndrome is essential if the diagnosis is to be confidently excluded. 
  • Morning or random serum cortisol is not suitable nor recommended by any guidelines for ruling out Cushing’s syndrome. 
  • The most useful tests are easily requested, readily arranged by the testing laboratory and generally convenient for the patient. Late night salivary cortisol, performed twice, is an excellent first choice.

Author competing interests – nil