All about dexamethasone testing in Cushing’s syndrome

Introduction

There are more than 60 years since Liddle first used dexamethasone, one of the most potent synthetic glucocorticoids, at a dose of 0.5 mg every 6 hours for 2 days, to significantly suppress glucocorticoid synthesis in normal subjects, mainly based on urinary metabolite assessment [ ]. Then, dexamethasone was given at 23.00 hours, and suppression was confirmed at 08.00 hours the next morning, as suggested by Nugent et al. [ ] and confirmed by Pavlatos et al. [ ]. Nowadays, dexamethasone tests are invaluable diagnostic tools in evaluating endogenous Cushing’s syndrome. They are used for both screening when suspicion of hypercortisolemia arises, and to help clarify the specific etiology. Dexamethasone is a potent synthetic corticosteroid, approximately 25-times more potent than cortisol, with a long duration of action [ ]. Most importantly, it does not interfere with the assays used to measure cortisol [ ]. The rationale behind dexamethasone tests is that the administration of exogenous steroids in normal individuals results in suppression of the hypothalamus–pituitary–adrenal (HPA) axis, while in patients with Cushing’s syndrome, the HPA axis is resistant to steroid feedback and this effect is not observed. Furthermore, while patients with pituitary-dependent Cushing’s syndrome, Cushing’s disease, may be partially responsive to dexamethasone, this is not usually the case with ectopic sources of ACTH.

Dexamethasone diagnostic tests in Cushing’s syndrome diagnosis

When Cushing’s syndrome is suspected, an initial diagnostic workup should be promptly performed. The suspicion will generally be based on clinical symptoms and signs, but it should be emphasized that exogenous sources of corticosteroids (oral, injected, inhaled, topical) must be excluded as far as possible. Two basic low-dose dexamethasone tests are recommended as core screening tests: (1) the 2 mg/day (0.5 mg/6 hourly) for 2 days low-dose dexamethasone suppression test (LDDST) and (2) the 1 mg overnight dexamethasone suppression test (ODST). Measurement is based on serum cortisol sampled at 09.00 hours at the end of the test. In a meta-analysis, dexamethasone tests have demonstrated the highest sensitivity of 98.6%, albeit with the lowest specificity (90.6%), in comparison with other tests [ ]. However, a reanalysis incorporating additional data on cortisol assays used in individual studies showed a median sensitivity and specificity of 100.0% and 90.9%, respectively. Most studies, approximately two-thirds, used a prespecified cutoff of 50 nmol/L (1.8 μg/dL) to enhance sensitivity, as proposed by the Endocrine Society Practice Guideline [ ]. Of note, sensitivities were high in a narrow range regardless of which specific assay was used, while specificities were somewhat lower over a wider range [ ]. In a comprehensive review, both the LDDST and ODST seemed to have comparable sensitivities (98%–100%) when the criterion of a posttest cortisol <50 nmol/L is applied. However, the 2-day test shows greater specificity (95%–100%) compared with the overnight test (88%) [ ]. Generally, the overnight test is the preferred preliminary screening tool due to its ease and convenience, and it is most useful in an outpatient setting to essentially exclude Cushing’s syndrome. The 48-hour test may be an alternative to the overnight test, or can be used in cases when the overnight test results are equivocal. It is also useful as a confirmatory test when the ODST is positive as it has a higher specificity. According to some, it is more accurate if conditions relating to pseudo-Cushing states are present [ ]. Furthermore, it may be important to supervise the administration of any tablets to ensure the protocol is followed in detail. When an overnight test protocol is followed, 1 mg of dexamethasone is usually taken between 23.00 and 24.00 hours and midnight, with subsequent blood collection for cortisol measurement the next day between 08.00 and 09.00 hours. Variations of the test with different doses of dexamethasone administration have been proposed, but without demonstrating any clear superiority with one another after careful comparison [ ]. Thus, 1 mg at midnight remains the standardized approach. It is of note that even the recent guidelines for the diagnosis of Cushing’s syndrome in the context of an adrenal incidentaloma still suggest the ODST to differentiate a nonfunctioning lesion from autonomous cortisol secretion [ ]. In terms of the LDDST, it is important that the patient is given instructions to take the dexamethasone precisely 6-hourly (i.e., 09.00, 15.00, 21.00, 03.00, 09.00, 15.00, 21.00, and 03.00 hours), ideally as written instructions.

Once the diagnosis of Cushing’s syndrome is confirmed, plasma ACTH levels should be checked to further categorize the etiology as ACTH-dependent or ACTH-independent. With an adrenal source, ACTH-independent Cushing’s syndrome, the serum cortisol will show little or no fall, and the next investigation would be scanning of the adrenals. In the case of ACTH-dependent Cushing’s syndrome, localization is then paramount, as the cause could be either pituitary (Cushing’s disease) or ectopic.

The high-dose dexamethasone suppression test (HDDST) was originally proposed to differentiate the source of ACTH overproduction (pituitary vs. ectopic). The rationale was somewhat similar to the low-dose DST, in that the administration of supraphysiological exogenous steroids could still evoke some degree of negative feedback on corticotroph tumor cells, albeit somewhat reduced on these neoplastic corticotroph cells, unlike ectopic ACTH-producing tumors, which would be more likely to be completely autonomous. There are two major protocols for the HDDST, performed as either 2 mg/6 hourly for 48 hours (as in the LDDST but replacing the 0.5 mg dose with 2.0 mg dexamethasone) or as an overnight 8 mg single-dose between 23.00 hours and midnight. Classically, during the 2-day test, urine collections for UFC (24-hour urinary free cortisol) are collected simultaneously with dexamethasone administration. Blood for cortisol measurement should be collected at 09.00 hours on day 2 and between 09.00 hours the next day, respectively. The proposed cutoff point for a positive response, suggestive of Cushing’s disease, is a decrease in basal serum cortisol or UFC by 50% or more [ ]. However, up to 30% of patients with Cushing’s disease show false-negative results to the test by not suppressing cortisol levels under the proposed cutoff, leading to potential misdiagnosis. Additionally, high false-positive rates and results (20%–30%) are observed in ectopic ACTH, especially as the traditional ectopic source, a small-cell lung cancer, is less often seen, and the more usual tumors (bronchial carcinoids, pancreatic neuroendocrine tumors, etc.) may show some degree of expression of normal corticotroph characteristics. The significant overlap in responses of Cushing’s disease and ectopic ACTH-producing tumors to HDDST compromises the diagnostic accuracy of this test [ ]. The sensitivity and specificity of the HDDST depend on the cutoff considered. A cutoff of plasma cortisol suppression of >50% yields a sensitivity of 69%–95% and a specificity of 60%–80%. Shifting the criteria can only increase the specificity at the expense of sensitivity, and vice versa [ ]. Therefore, we have suggested that the HDDST has now outlived its usefulness. Others have suggested it may have a place in the diagnostic algorithm if bilateral inferior petrosal sinus sampling and pituitary MRI scanning are not available, and then only in conjunction with a corticotrophin-releasing hormone (CRH) stimulation test, but unfortunately CRH is no longer available [ ]. The combination of the CRH test and the HDDST is indicative of CD when a positive response is observed. Approximately 90% of patiens with ectopic ACTH-producing tumors, but only 5% of those with CD, failed to respond to either the CRH test or the HDDST [ ]. In two studies, if the CRH stimulation test and the overnight 8 mg DST were analyzed in combination, sensitivities (of 76% and 81%) and specificities (of 89% and 100%) were reported [ , ]. Interestingly, Isidori et al. showed that more than 30% suppression during the LDDST almost invariably predicted full suppression (>50%) during the HDDST, and the combination of results from the individual LDDST and CRH tests achieved a diagnostic accuracy similar to HDDST and CRH, exhibiting a sensitivity of 94% and a specificity of 97%, respectively [ ].

An outpatient procedure utilizing intravenous dexamethasone has been suggested to overcome potential concerns about drug compliance and absorption of oral dexamethasone. Various dosages of infusion ranging from 1.75 to 10.5 mg (4 mg most common) and a single bolus of 8 mg have been evaluated. Serum cortisol is checked at regular intervals and then after 22–24 hours. A day-2 serum cortisol level, which is greater than 20% of the baseline value [or greater than 130 nmol/L (4.7 μg/dL)], is indicative of a Cushing’s syndrome diagnosis with a sensitivity of 100% and specificity of 96%. Using >200 nmol/L (7.25 μg/dL) as a cutoff, the sensitivity was 98% with a specificity of 100%. It is also helpful in distinguishing the etiology of Cushing’s syndrome: in Cushing’s disease at the end of infusion, serum cortisol levels usually show suppression from the baseline, followed by rebound hypercortisolism at 24 hours. However, ectopic ACTH-producing tumors and adrenal Cushing’s syndrome do not show a similar pattern as cortisol levels are not suppressed. Measuring cortisol levels on day 1 (+5 hours) rather than +24 hours is pivotal for distinguishing these conditions, with >70% suppression indicating Cushing’s Disease. Nonetheless, few studies have used this procedure to distinguish the source of ACTH–dependent Cushing’s syndrome due to the limited number of patients with an ectopic source [ ]. It has also been used in a small number of patients to differentiate Cushing’s disease from functional hypercortisolism, with a sensitivity and specificity of 90% and 80%, respectively [ ].

The sequential dexamethasone-CRH suppression test (dex-CRH test) is based on the premise that the suppression HPA axis by the administration of exogenous steroids can be overcome by CRH stimulation at the pituitary level in normal subjects. This test has been used to differentiate patients with CD based on mild signs and symptoms, and a positive screening test, from functional hypercortisolemia (pseudo-Cushing). Yanofski et al. first introduced the dexamethasone-CRH test in 1993: 53 patients with Cushing’s syndrome, 19 of whom were diagnosed with pseudo-Cushing’s syndrome, were evaluated, and the dexamethasone-CRH test showed a greater diagnostic accuracy than either the LDDST or CRH test alone in the diagnosis of Cushing’s syndrome [ ]. Dexamethasone 0.5 mg is given every 6 hours for 48 hours, and then intravenous ovine-sequence CRH (1 μg/kg) is administered 2 hours after the last dexamethasone dose. The serum cortisol level is measured 15 minutes later. However, the additional CRH stimulation to the LDDST seemed to improve the sensitivity (100%) at the expense of specificity (95%) [ ]. Different cutoffs of cortisol levels yielded different results. As such, using a cutoff >38 nmol/L (1.4 μg/dL) the dex-CRH test showed sensitivity ranging between 90% and 100% and specificity ranging between 50% and 100%, respectively. However, when the >87 nmol/L (>3.2 μg/dL) cutoff was applied, sensitivity and specificity were 94% and 100%, respectively. The test is cumbersome and requires an inpatient stay, whereas it has not been validated in many centers and, most critically, due to the lack of availability of CRH, it is clearly no longer in use [ ] ( Table 11.1 ).

Table 11.1

Dexamethasone diagnostic tests (DST) used for the establishment of hypercortisolemia.

Dexamethasone suppression tests
Types of DST		Sensitivity	Specificity	Indications
Low dose DST	2 mg/day for 2 days (LDDST)	98%–100%	97%–100%	Diagnosis of endogenous CS-LDDST distinguishes also pseudo-CS [ ]
Low dose DST	Overnight 1 mg (ODST)	98%–100%	87.5%
High dose DST	8 mg/day for 2 days	69%–95%	60%–80%	Etiology of ACTH-dependent CS (CD or ectopic) [ ]
High dose DST	Overnight 8 mg	69%–95%	60%–80%	Etiology of ACTH-dependent CS (CD or ectopic) [ ]
Combination HDDST and CRH	8 mg overnight dexamethasone or 2 mg/6 hourly for 2 days and bolus CRH 100 μg	76%–95%	89%–100%	Etiology of ACTH-dependent CS (CD or ectopic) [ ]
Combination LDDST and CRH	0.5 mg/6 hourly dexamethasone for 2 days and bolus CRH 100 μg	94%	97%	Etiology of ACTH-dependent CS (CD or ectopic) [ ]
Dex-CRH test	0.5 mg/6 hourly for 48 hours dexamethasone and CRH (1 μg/kg) 2 hours after the last dexamethasone administration	90%–100% (>38 nmol/L) 94% (>87 nmol/L)	50%–100% (>38 nmol/L) 100% (>87 nmol/L)	Differentiate between CD and pseudo-CS [ ]
Intravenous dexamethasone	Infusion of 1.75–10.5 mg or an 8 mg bolus	100% (>20% from baseline or 130 nmol/L on day 2)	96% (>20% from baseline or >130 nmol/L on day)	Diagnosis of endogenous CS [ ]
		90%	80%	Differentiate CD from pseudo-CS [ ]
		Limited data	Limited data	Etiology of CS [ ]