Glucocorticoid Therapy and Cushing Syndrome Workup

Updated: Mar 23, 2021
  • Author: George P Chrousos, MD, FAAP, MACP, MACE, FRCP(London); Chief Editor: Sasigarn A Bowden, MD, FAAP  more...
  • Print

Approach Considerations

In the investigation of patients with hypercortisolism, the first task is to determine whether the source is endogenous because the overwhelming majority of patients with Cushing syndrome (CS) take glucocorticoids for therapeutic or other purposes or, in rare cases, surreptitiously. If no evidence suggests exogenous glucocorticoid administration, the next task is to establish that cortisol levels are indeed elevated before determining the underlying cause. The following is a diagnostic approach to the patient with Cushing syndrome.


Laboratory Studies

Tests to establish the presence of Cushing syndrome

Twenty-four–hour urinary free cortisol (UFC) assay

Reference range values are less than 70 mg/1.73 m2/24 h. Reference ranges vary, depending on the assay method and the antibody specificity. UFC values must be interpreted in light of the reference range for the local laboratory. Results must be corrected for body surface area in children.

Comeasurement of creatinine may help detect collections that are incomplete or mistimed because 24-hour creatinine excretion should be similar for each of the 3 collections. Perform a minimum of 3 collections if initial measurements are low because 5-10% of patients with Cushing syndrome have an intermittent or periodic pattern of excretion.

Twenty-four–hour urine 17-hydroxy (17-OH) steroids test

This test includes all cortisol metabolites with the 17-(OH)2-acetone side chain. 17-OH steroids are not routinely used because they are difficult to assay, and UFC measurements are equally reliable. Reference range values are less than 6 mcg 17-OH steroids per gram of creatinine.

Low-dose dexamethasone suppression test

Administer 15 mcg/kg dexamethasone at 11 pm. Patients with Cushing syndrome do not suppress their cortisol to less than 3 mcg/dL at 8 am the following morning. This test is associated with a less than 3% false-negative rate and a 20-30% false-positive rate.

Patients with psychiatric disorders that cause corticotropin-releasing hormone (CRH) overactivity (eg, anxiety, depression, posttraumatic stress disorder) may not suppress with low-dose dexamethasone.

Diurnal serum cortisol measurement

Patients with Cushing syndrome who have hypercortisolemia at the time of assessment have loss of the normal nocturnal fall in cortisol. Measurements should be taken at 11:30 pm and midnight and then at 7:30 am and 8:00 am through a cannula that has been inserted 1-2 hours before and with the patient fasting since the evening meal. A single cortisol measurement at midnight through an indwelling cannula and with no food since the evening meal is equally sensitive. The result should be less than 5 mcg/dL. Patients with intermittent or periodic disease may have normal diurnal variation, if disease is quiescent at the time of assessment. This test is very sensitive; the main disadvantage is the requirement of overnight admission.

Tests to establish the cause of Cushing syndrome

Corticotropin (ACTH) measurements

Correct collection of ACTH is essential. The ACTH peptide is unstable and should be collected in ethylenediaminetetraacetic acid (EDTA)–containing tubes on ice, spun soon after collection, and stored at -20°C until assay.

Comeasurement of cortisol and ACTH can distinguish ACTH-independent Cushing syndrome from ACTH-dependent Cushing syndrome because ACTH is suppressed in the former if hypercortisolemia is present at the time of measurement. Absence of full suppression of ACTH does not exclude ACTH-independent Cushing syndrome because, in mild disease or in periodic Cushing syndrome, ACTH responsiveness may persist, particularly if the disease process is quiescent when sampling occurs.

Random measurements of ACTH are of limited value, although elevation of paired ACTH and cortisol measurements in the evening is suggestive of Cushing disease.

Patients with Cushing syndrome due to ectopic ACTH secretion typically present earlier with more florid symptoms, and associated hypokalemia may be present.

Ectopic ACTH production may lack the normal processing so that ACTH-precursor molecule levels are commonly elevated. Radioimmunoassays (RIAs) are the least specific and may detect a greater proportion of precursor molecules than the more specific double-antibody immunoradiometric assays (IRMAs) or chemiluminescent assays. Assay by more than one technique may help to demonstrate disparate values, indicating the presence of incompletely processed molecules, as observed in ectopic tumors.

CRH stimulation test (ovine or human CRH)

The purpose is to distinguish Cushing syndrome from ectopic ACTH secretion.

This test is performed in the morning, after an overnight fast. Measure ACTH and cortisol at -5 minutes, at baseline, and at 15-minute intervals for 90 minutes. Administer 1 mcg/kg of ovine or human CRH intravenously.

A positive response to this test can also be observed in healthy people; consequently, confirming hypercortisolemia before interpreting the results is important.

A rise in cortisol of 20% or more at 15 minutes and 30 minutes and cortisol levels above baseline at -5 minutes and 0 minutes is 91% sensitive and 88% specific. A rise in the mean of the ACTH values of greater than 35% above baseline levels is 91% sensitive and more than 99% specific in diagnosing Cushing syndrome.

Patients with ectopic ACTH secretion tend to have a rise in ACTH of less than 35%, although the probability of Cushing disease remains high at all levels of responses, which suggests that, in the absence of a discrete lesion on pituitary imaging, performing a second test (ie, high-dose dexamethasone suppression test, bilateral inferior petrosal sinus sampling [BIPSS]) to confirm the diagnosis is prudent.

High-dose overnight dexamethasone suppression test

This test has a role in differentiation of ACTH-dependent Cushing syndrome due to a pituitary adenoma from ectopic ACTH production.

On the day that dexamethasone is to be administered, measure the cortisol level at 8 am. The dexamethasone dose of 120 mcg/kg, maximum dose 8 mg, is administered at 11 pm, and a further cortisol measurement is obtained at 8 am the following morning.

Cortisol suppression of greater than 50% has a sensitivity that is comparable with the 85% sensitivity of the formal Liddle test for diagnosing Cushing syndrome.

Some ectopic sources of ACTH production produce false-positive results, and some large ACTH-producing macroadenomas may not suppress with high-dose dexamethasone. This test has the advantages of being inexpensive and requiring no admission or timed urine collections. Interindividual variation in the pharmacodynamics of dexamethasone metabolism is significant, which must be considered if the test results are equivocal or at odds with other test results. This problem can be overcome by measuring dexamethasone levels at the time of the second cortisol measurement, although the assay is not widely available.

Other tests used in the investigation of Cushing syndrome

Metyrapone test

The metyrapone test can also be used to differentiate ACTH-secreting pituitary adenomas from ectopic ACTH-secreting tumors. Metyrapone blocks the adrenal enzyme 11-hydroxylase, the most prominent place of inhibition being the conversion of 11-deoxycortisol to cortisol.

The classic test entails the administration of oral metyrapone with food at 4-hour intervals for 24 hours, with measurement of plasma 11-deoxycortisol concentrations before and 4 hours after the last dose. Cortisol is also measured 4 hours after the last dose to confirm adequacy of the blockade of 11-hydroxylase. In patients with Cushing syndrome, serum 11-deoxycortisol levels increase to greater than 5 mg/dL (144 nmol/L) in the presence of a low or undetectable cortisol level.

Concomitant measurement of 24-hour 17-OH steroid measurements (cortisol degradation products) improves the sensitivity to 60%, with a specificity of 77%.

Twenty-four–hour 17-OH steroids should rise by greater than 70% over baseline. Levels are measured on the day before, during, and after metyrapone administration. The resulting fall in cortisol levels triggers an increase in ACTH by both the normal pituitary and ACTH-secreting pituitary adenomas but not by ectopic sources of ACTH.

Liddle dexamethasone suppression test

The Liddle test has been used to differentiate Cushing syndrome due to ACTH-secreting pituitary tumors from ectopic ACTH production and ACTH-independent Cushing syndrome. Grant Liddle developed this test in the late 1950s, before reliable ACTH assays were available. With the advent of reliable assays for ACTH and the demonstration that the Liddle test is of equal sensitivity to the high-dose overnight dexamethasone suppression test, it is used less frequently today. The main reason for the less frequent use of this test is that it requires an inpatient stay and takes 6 days to perform.

Dexamethasone is administered at a dose of 7.5 mg/kg every 6 hours for 2 days (low dose), then 30 mg/kg every 6 hours for 2 days (high dose). Twenty-four–hour urine collections for UFC and 17-OH steroids are performed for 2 days before the test and then during the test.

Patients with pseudo-Cushing states suppress with low-dose dexamethasone. In patients with Cushing disease, the abnormal corticotrophs are sensitive to glucocorticoid inhibition only at the high dose of dexamethasone. Patients with the ectopic ACTH syndrome or cortisol-secreting adrenal tumors usually do not respond even to high doses. The criterion for a positive response consistent with Cushing disease is a greater than 50% fall in 17-OH excretion on day 2 of high-dose dexamethasone (80% diagnostic accuracy). However, diagnostic accuracy of the test increases to 86% by measuring excretion of both UFC and 17-OH and by requiring greater suppression of both steroids (>64% and 90%, respectively, for 100% specificity).

Dexamethasone-suppressed CRH test

This is used to distinguish Cushing disease from pseudo-Cushing disease. Reserve this test for patients with borderline or mild hypercortisolemia or for patients in whom cortisol suppression did not occur following low-dose (ie, 15 mcg/kg) overnight dexamethasone test, when the diagnosis of Cushing disease is uncertain.

Administer 7.5 mcg/kg of dexamethasone, not to exceed 0.5 mg, every 6 hours for 48 hours starting at noon and finishing at 6 am. At 8 am on the morning that the dexamethasone is completed, CRH (1 mcg/kg) is administered intravenously and cortisol and ACTH are measured at -10, 0, +5, and +15 minutes.

A 15-minute cortisol level that exceeds 1.5 mcg/dL (38 nmol/L) is diagnostic of Cushing disease. In pseudo-Cushing disease, dexamethasone results in corticotroph suppression by negative feedback on the pituitary, hence failure to respond to CRH. The dexamethasone-CRH test achieves nearly 100% specificity, sensitivity, and diagnostic accuracy.

Thyroid function tests

Thyroid function tests and measurement of insulinlike growth factor-1 (IGF-1) and insulinlike growth factor–binding protein-3 (IGF-BP3) are effective to screen for hypothyroidism and GH deficiency. Patients with a low IGF-1 or IGF BP3 must have further testing to confirm the diagnosis of growth hormone (GH) deficiency and establish the cause.

Establishment of parathyroid hormone (PTH) resistance

Establishment of the presence of PTH resistance (low calcium, elevated phosphate, elevated PTH with reference range 25-OH vitamin D, and measurement of urinary cyclic adenosine monophosphate [cAMP] response to PTH) with or without thyroid-stimulating hormone (TSH) resistance can be used to identify patients with pseudohypoparathyroidism.


Imaging Studies

Choose imaging studies as directed by the results of the clinical assessment and the initial biochemical investigations. Always obtain imaging studies of the pituitary if the results indicate that the patient has an ACTH-dependent cause of Cushing syndrome, even if the results suggest a possible ectopic cause, because very large ACTH-secreting pituitary adenomas behave more like ectopic ACTH-producing tumors.

Investigations that should be performed both in patients with hypercortisolism and in those with proven Cushing syndrome are as follows:

Bone age

Perform bone age radiography as part of the assessment of the child with short stature. Children receiving pharmacologic doses of glucocorticoids may have a delayed bone age either because of their treatment or because of their primary disease. On the other hand, a child with Cushing syndrome may have a bone age that is only mildly delayed or even equal to the chronologic age because of the coexistence of hyperandrogenemia. In the latter case, this has a significant adverse impact on the child's final height.

Assessment of bone mineral density

Assessment of bone mineral density (BMD) is important whenever hypercortisolism has been present for a significant length of time. In patients receiving pharmacologic doses of glucocorticoids, assess at baseline and at 12-month intervals thereafter while the patient is receiving steroid treatment. [10]

In patients with Cushing syndrome, perform BMD measurement once the diagnosis has been established. Loss of bone density is a sensitive marker of hypercortisolism; therefore, BMD may be helpful in cases where the diagnosis is uncertain and cortisol levels are elevated only mildly.

When assessing BMD in children, use a technique that has established reference range values that are age and sex appropriate, ensuring the necessary volumetric correction is made for the size of the child.

Imaging methods for children suspected of having Cushing disease are outlined below.

Pituitary MRI

MRI is the modality of choice for imaging the pituitary gland. This technique is preferred to CT scanning because of its superior resolution. Moreover, image quality is not lost because of the surrounding bone of the skull base, and images can be obtained in multiple planes. MRI should include sagittal and coronal images taken through the pituitary and parasellar region at 2-mm to 3-mm intervals following intravenous gadolinium. Pituitary adenomas typically appear as round or oval hypoenhancing lesions. Relative hypoenhancement is due to enhancement of the surrounding normal gland with gadolinium and delayed uptake in the tumor.

False-positive results can occur because of the phenomenon of signal averaging, or excessive noise, if cuts are too thin. False-negative results may occur if the time taken between gadolinium injection and imaging is too long because gadolinium uptake by the tumor eventually occurs. To be confident of the result, the tumor should ideally be visible on at least 2 images. Otherwise, or if other results are equivocal, the patient should undergo inferior petrosal sinus sampling.

CT scanning

CT scanning is not recommended because the image resolution may not allow detection of microadenomas. Refer the patient to a center with a powerful MRI scanner and with personnel experienced in performing and interpreting scans.

Adrenal imaging

Contrast enhanced CT scanning is the modality of choice for imaging the adrenal glands. Because adrenal adenomas are frequently small and may be missed, this technique is superior to ultrasonography. CT scanning can readily distinguish adrenal glands that are high in fat content from adjacent liver, spleen, and kidney, which are of lower attenuation. If significant insulin resistance is present, the liver may appear brighter than normal on an unenhanced scan, suggesting fatty infiltration.

Cortisol-secreting adrenal adenomas lead to suppression of pituitary ACTH production, with resultant atrophy of the contralateral adrenal gland and the remainder of the ipsilateral gland.

Consider any adrenal mass in a child suspicious because incidental tumors are rare. The presence of bilateral hyperplasia or nodularity of the adrenal glands should raise suspicion of an ACTH-dependent process or a congenital abnormality of the adrenal gland that leads to hyperplasia and autonomous cortisol hypersecretion, such as McCune-Albright syndrome or Carney complex. Patients with McCune-Albright syndrome who develop Cushing syndrome usually present in infancy and have fairly extensive disease. Other manifestations of McCune-Albright syndrome include bony lesions (detectable on bone scanning or plain radiography), patchy skin pigmentation, and other endocrine manifestations. Modern CT scanners are able to resolve lesions as small as 5-10 mm in size. Although adrenal masses may develop hemorrhagic areas, calcification is unusual in benign cortical lesions; thus, calcification suggests the possibility of malignancy or of pheochromocytoma.

T2-weighted MRI is the modality of choice for suspected ACTH-secreting adrenal pheochromocytomas or malignant adrenocortical carcinomas.

Iodocholesterol scanning

Iodocholesterol scanning can be used to identify adrenal cortical tumors, although it has no benefit over thin section CT scanning with a modern CT scanner, and it is mainly used as a second-line imaging modality. This test is useful in the localization of ectopic adrenocortical tissue or adrenocortical tissue that was not removed in a bilateral adrenalectomy.

Ectopic-ACTH secretion studies

Imaging studies performed in cases of suspected ectopic-ACTH secretion are outlined as follows:

Small cell carcinoma of the lung is the most common cause of ectopic ACTH secretion in adults. Ectopic ACTH secretion is very rare in children, commonly arising from a carcinoid tumor in the chest or abdomen, although it may arise from a neuroendocrine tumor in the pancreas (especially multiple endocrine neoplasia [MEN1]) or rarely from ganglioneuromas and pheochromocytomas of the adrenal medulla. Perform CT scanning of the chest (ie, for bronchial carcinoid, thymic neoplasms, possible mediastinal metastases) and the abdomen with intravenous and gastrointestinal contrast, focusing on the pancreas, the liver, the duodenum, and the appendix. MRI may detect lesions missed by CT scanning and should be used adjunctively in cases where ectopic ACTH secretion is suspected.

ACTH-producing neuroendocrine tumors may also be detected using radioisotope scanning. Octreotide scanning and fluoro-dopamine positron emission tomographic (PET) scanning may aid in the identification of small neuroendocrine tumors. These techniques are still investigational and are not routinely recommended in children.



The two procedures required in the workup of a patient with Cushing syndrome are BIPSS and bilateral adrenal vein sampling. These procedures both require a high level of technical expertise and should only be performed when indicated in centers that perform such sampling regularly.

Complications may arise from the invasive testing that some patients with Cushing syndrome require. Adrenal vein sampling can be associated with hemorrhage into the adrenal gland, which, if it occurs in the unaffected side, may render the patient permanently adrenally insufficient, following excision of their adenoma. Inferior petrosal sinus sampling can rarely cause bleeding or thrombosis of petrosal sinuses with neurologic sequelae. Therefore, reserve these procedures for occasions when findings from prior investigations are either contradictory or inconclusive.

BIPSS with injection of CRH

Venous blood from the anterior pituitary drains into the cavernous sinus and subsequently into the superior and inferior petrosal sinuses.

This is performed to differentiate Cushing disease from ectopic ACTH-secreting tumors.

BIPSS involves the simultaneous sampling of ACTH from each inferior petrosal sinus and from a peripheral vein before and after the injection of CRH. Catheters are advanced into both inferior petrosal sinuses via the ipsilateral femoral veins. Samples are simultaneously collected for ACTH from each inferior petrosal sinus and a peripheral vein before and after (at 2 min, 5 min, and 10 min) injection of 1 mcg/kg intravenous of ovine or human CRH.

Patients with the ectopic ACTH syndrome have no ACTH concentration gradient either between the inferior petrosal sinuses or between the central and peripheral samples. A ratio greater than or equal to 2 in basal ACTH samples between either or both of the inferior petrosal sinuses and a peripheral vein is highly suggestive of Cushing disease (95% sensitivity, 100% specificity).

Stimulation with CRH during the procedure, with the resulting outpouring of ACTH, increases the sensitivity of BIPSS for detecting corticotroph adenomas to 100% when peak central to peripheral ACTH ratio is greater than or equal to 3.

Petrosal sinus sampling must be performed bilaterally and simultaneously because the sensitivity of the test drops to less than 70% with unilateral catheterization. BIPSS is technically difficult and, like all invasive procedures, can never be risk free, even in the most experienced hands. Reserve BIPSS only for patients with possible Cushing disease and negative or equivocal findings on MRI of the pituitary and for patients with positive findings on pituitary MRI but equivocal findings on suppression and stimulation tests. In the former group, BIPSS unequivocally distinguishes ACTH-secreting pituitary adenomas from ectopic sources of ACTH production, and it may provide lateralization data of potential value to the surgeon. In the latter group, BIPSS findings exclude the possibility of a pituitary incidentaloma (rare in children but present in as many as 10% of adults).

Lateralization of pituitary adenomas with BIPSS assumes that blood flow is equal on each side. As a result, if vascular drainage is aberrant or if the sinuses are not of equal size, false-positive lateralization may occur. Similar to other biochemical tests for Cushing syndrome, the results of the BIPSS are valid only when the patient is hypercortisolemic.

Bilateral adrenal vein sampling

Cortisol is synthesized only in adrenal cortical tissue. Perform sampling in a patient with apparent ACTH-independent Cushing syndrome when the tumor is not identified on imaging studies or when the source of cortisol excess (ie, unilateral or bilateral) is not clear.

Sensitivity and specificity are optimized when simultaneous sampling of both adrenal veins and a peripheral vein is performed after stimulation with Cortrosyn. Adrenal venous cortisol levels usually should be greater than peripheral levels. Marked elevation of cortisol in one adrenal vein with reference range or suppressed levels in the other is suggestive of a neoplasm on the side with elevation. Levels in the suppressed side are still likely to be higher than peripheral levels. If they are the same as peripheral levels, then consider the possibility that the cannula has slipped out of the vein (this is particularly likely on the right side).

This procedure has a significant risk of morbidity from hemorrhage into the adrenal if the catheter is flushed.


Histologic Findings

Pituitary adenomas show evidence of acinar expansion, clonal expansion of a population of cells immunopositive for ACTH. Loss of the normal reticulin pattern is evident. The Crooke hyaline change may also be observed because of the hypercortisolism.

The histologic appearance of the tumor is similar in Nelson syndrome, except that the latter is more likely to show evidence of nuclear and cellular pleomorphism. Pituitary malignancies are extremely rare. Adrenal cortical neoplasms may be benign or malignant. Malignant tumors usually have evidence of nuclear pleomorphism. Mitoses are uncommon, and their presence with vascular invasion is diagnostic of malignancy.

In the absence of distant metastases, correlation between histologic appearance and tumor behavior is not predictable. A correlation between tumor size at diagnosis and risk of malignancy is observed. Tumors less than 100 g are more likely to be benign.