Final Diagnosis -- Cushing's Disease: Hypercortisolism and Pituitary Adenoma



Clinical Presentation:

Cushing's syndrome (CS) consists of a cluster of symptoms related to excessive exposure to glucocorticoids, of either endogenous or exogenous origin. Cushing's disease refers to the pituitary-dependent Cushing's syndrome (1). Classic clinical features include central obesity and rounded face, 'buffalo hump' caused by fat deposit on the posterior neck, purple skin striae, and thinned extremities due to muscle atrophy. Images of these physical findings can be seen at Hirsutism and alopecia, nephrolithiasis (often subclinical), osteoporosis with pathologic fracture, gonadal dysfunction, neurologic difficulties including impairment in cognition and memory, and psychiatric symptoms ranging from anxiety or depression to psychosis also occur (1). However, patients may present with only isolated signs (2).

Endogenous Cushing's syndrome may be either corticotropin (ACTH)-dependent or -independent (5). ACTH-dependent etiologies comprise 80% of all cases and are associated with bilateral adrenocortical hyperplasia and include Cushing's disease, ectopic secretion of corticotropin by non-pituitary tumors, and ectopic secretion of corticotropin-releasing hormone (CRH) by non-hypothalamic tumors (3). Cushing's disease secondary to a pituitary adenoma is the most common, accounting for 70% of ACTH-dependent cases (10). ACTH-independent cases are related to exogenous administration of glucocorticoids or to primary adrenal diseases including adrenocortical adenomas or carcinomas, macronodular bilateral adrenal hyperplasia, primary pigmented nodular adrenal disease, and McCune-Albright syndrome. The latter three account for only a small number of cases (4).


In this case, the patient presented first with radiographic findings. Upon discovery of the pituitary lesion, its functionality had to be determined. Laboratory evaluation of the various pituitary hormones comprised her initial evaluation. ACTH and cortisol were the only hormones present in abnormal levels, and so the clinicians proceeded to confirm the diagnosis of ACTH-dependent hypercortisolism.

The evaluation of possible Cushing's syndrome is a complicated one. Various tests are available; none have sufficient sensitivity and specificity to be used separately. Furthermore, multiple atypical clinical presentations are possible and include mild hypercortisolism, cyclical CS with spells of hypercortisolism alternated with normal secretion, subclinical CS, and pseudo-CS related to pregnancy or alcoholism (2, 10).

Usually, the initial diagnostic goal is to confirm CS by documenting hypercortisolism. Cortisol is secreted intermittently, with peak values occurring in the morning and lower levels in the evening (11). Therefore, an initial evaluation often includes a morning cortisol, with reference range at UPMC of 7-25 ug/dL. However, patients with CS frequently have normal morning concentrations (1). First, other laboratory methods utilized in this patient's case will be discussed, followed by additional possible options.

Nighttime Salivary Cortisol: Salivary cortisol concentration correlates highly with free plasma cortisol, is not affected by circulating levels of cortisol binding globulin, and is independent of salivary flow rates (2). Sample collection, while simple, should adhere strictly to laboratory instructions regarding collection device and sample collection duration(4). Specimens are collected between 11:00 PM and midnight by having the patient chew on a cotton swab for a laboratory-specified period of time store the swab in a plastic tube in the refrigerator before bringing it to the laboratory. Salivary cortisol concentrations fluctuate throughout the day with lowest level at night-time. The reference range will vary somewhat by laboratory and by analytical method. A patient with CS will have an elevated late night salivary cortisol level (2).

Twenty-Four Hour Urinary Free Cortisol (UFC): This test offers a measure of free cortisol secreted over an extended period of time. Also, these results are unaffected by factors that influence corticosteroid binding globulin (CBG) levels and the fraction of bound versus unbound hormone (1). Values greater than four times the upper limit of normal are very unusual except in CS and can be considered diagnostic (1). If the clinical suspicion is high and the initial result is normal, the test can be repeated up to three times, as the hypercortisolism may be intermittent. Determination by immunoassay can be affected by varying degrees by steroid metabolites or synthetic glucocorticoids, specific for each method. HPLC has greater sensitivity and specificity; however, interfering substances (including carbamazepine and digoxin) do occur. Methods using mass spectroscopy combined with GC or HPLC has superior sensitivity and specificity, but may not be readily available except at reference laboratories (2).

Low Dose Dexamethasone Suppression Test (LD-DST): Nugent's overnight version of this test involves a single oral dose of 1mg dexamethasone between 11:00 p.m. and midnight followed by determination of fasting plasma cortisol at 8:00 the following morning (10). The newest recommendation for cut-off of normal cortisol suppression is less than 1.8ug/dl (50 nmol/L). Suppression below that level discounts active CS, although increased CBG, acute or chronic illness, or pseudo-CS can alter the results (2). Liddle's classic form of this evaluation, used for this patient, entails oral 0.5mg dexamethasone doses every 6 hours for two days. A 24-hour urine specimen is collected on the second day for UFC determination. Suppression of UFC to less than 10 ug per 24 hours or the following morning's fasting plasma cortisol to less than 1.8 ug/dL also excludes CS(3,10). The investigation can also be performed with an additional stimulus step with corticotropin releasing hormone (2).

After hypercortisolism has been established, the next diagnostic step is to distinguish amongst the three forms of CS as discussed above: exogenous, endogenous and ACTH-dependent, or endogenous and ACTH-independent (1).

Plasma ACTH: If plasma ACTH is suppressed to less than 5 pg/mL, adrenal-dependent ACTH-independent disease is favored and CT or MRI adrenal imaging is warranted. If plasma ACTH is greater than 20 pg/mL, as in this case, ACTH-dependent causes should be investigated (1). Although ACTH levels tend to be higher in ectopic production, there is significant overlap that prevents use of ACTH alone to distinguish between the conditions (2). Cases with intermediate ACTH values may require further confirmation by dynamic testing such as a high dose dexamethasone suppression test (1).

Pituitary Contrast-Enhanced Magnetic Resonance Imaging: In contrast to this case, radiographic features of pituitary anomalies are not commonly the presenting finding. Pituitary MRI with gadolinium enhancement is performed in most patients following confirmation of ACTH-dependent CS, and will identify a discrete lesion in up to 60% of patients (2). The most common form of active adenoma is a prolactinoma and these lesions usually appear on pre-contrast T1 weighted images as areas of abnormal low signal (see patient's MRI above). One review noted that the overall sensitivity of MRI for detecting prolactin-producing microadenomas, using various techniques, approaches 90%. However, the corticotrophin-secreting adenomas of CS seem to be more difficult to image by MR with sensitivity of only 50-60% (7). Another review confirmed that still almost 50% of adenomas will go unnoticed (3). However, if a focal lesion greater than 5mm in size is identified, no further work-up may be necessary. Although 10% of the population has a pituitary incidentaloma, most are smaller than 5mm (1). Also, adenomas were histologically confirmed more frequently in patients with a positive versus negative MRI, 40 and 81% respectively (12).

Bilateral Inferior Petrosal Sinus Sampling (IPSS): A dural venous sinus into which the pituitary hormones drain, the inferior petrosal sinus (IPS), extends from the posterior cavernous sinus to join the internal jugular vein. Bilateral inferior petrosal sinuses communicate through the basilar venous plexus and with the marginal sinus. Diagrammatic representations of this anatomy, and its known variations, can be found in Lad et al and Miller and Doppman (3, 8). In this technique, an interventional radiologist catheterizes both inferior petrosal sinuses from the internal jugular veins via sheaths in the femoral veins and serial blood samples are drawn from each sinus and a peripheral line and evaluated for ACTH over time, before and after stimulation with CRH. A gradient between central and peripheral ACTH values arises from elevated levels in venous drainage of the pituitary. If the ACTH is produced ectopically, no such gradient exists at the level of the petrosal sinuses. Prior to CRH administration, a basal ratio of central to peripheral ACTH of 2.0 or greater strongly suggests CD. Following CRH administration, a ratio greater than 3.0 is also diagnostic (1,3). Aberrant results could be due to anomalous sinus anatomy, intermittent or cyclical CS, or CRH-secreting tumors. (1)

Failure of central localization in IPSS should prompt a search for extra-pituitary sites of ACTH secretion, although a pituitary source cannot be entirely ruled out (1). Several factors can result in misleading results. Patient selection is particularly important. Patients should have documented hypercortisolism, must not be taking any suppressive drugs, and should not have already had bilateral adrenalectomy. Furthermore, variant IPS anatomy can make cannulation difficult. In fact, in 24% of patients in one study, the IPS existed as a plexus of veins, rather than a single vessel, that could not be reliably cannulated from the internal jugular vein (8). As the validity of the study depends upon accurate cannulation, venography should be performed for confirmation. Images of normal and aberrant inferior petrosal sinus anatomy by venography can be viewed at Sampling errors secondary to dilution of pituitary blood from the anastamoses with other dural sinuses can also cause misleading results (9).

For confirming a central ACTH source, one study found IPSS to have a diagnostic accuracy of 86% using the baseline ratio and 90% using the post-stimulation ratio, compared to 50% for MRI and 40% for CT (Table 3). However, false positive results have also been reported (9). Another study of 179 patients reported a false positive (test indicates central source, but ectopic source later found) rate of 1% and a false negative (test indicates against central source, but pituitary tumor later identified) rate of 5% (13).

The notion that IPSS may be useful in localization of adenomas within the gland is controversial (2). One review claims that location was correctly predicted in 78% of cases with a side-side gradient of 1.4 or greater (1). The Colao et al study found a diagnostic ACTH inter-sinus ratio > 1.4 in 58% of patients. Basal and post-CRH-stimulus inter-sinus gradient concordance was 87%. Furthermore, the site of adenoma as predicted by IPSS, MRI, and CI concurred with surgical evidence in 65%, 75%, and 79% of cases, respectively (9). However, reversal of the inter-sinus ratio between pre- and post- stimulus values has been reported, and occurred in this case. This underscores the fact that IPSS is less reliable than MR or CT for adenoma localization (9). Additionally, histological evidence of a corticotroph pituitary adenoma is lacking in nearly 15% of patients with positive IPSS results (1).

Alternative sampling methods, from the jugular veins or cavernous sinuses, have also been suggested. Jugular venous sampling would be simpler from the standpoint of technique. One paper mentions the sensitivity of IPSS to be 94% compared with 83% for jugular sampling. ACTH values from the jugular vein are usually lower than those from the IPSS due to dilution within the large vessel. On the other hand, the cavernous sinuses are anatomically closer to the pituitary gland, and sampling here may be less subject to dilution. Indeed, this method accurately localized the lesion in 73% without CRH stimulation and 93.3% with stimulation. This method is not commonly used, however, given its increased invasiveness and greater risk of neurologic complications (3).

Although rare at experienced centers, serious complications such as deep venous thrombosis, pulmonary emboli, cranial nerve palsies, and brainstem vascular damage can occur (1,4). This evaluation should be considered in patients with CD whose previous studies have been discordant or ambiguous (1).

Other Possible Diagnostic Techniques:

Cortisol Circadian Rhythm / Midnight Cortisol: Disruption in the circadian rhythm of cortisol is an important feature of CS (4). A patient with CS may have a morning cortisol value within or only slightly greater than the normal range, but he or she will lack the normal fluctuation and have an elevated midnight plasma cortisol above 1.8 ug/dL (1). However, use of this method is limited as the venipuncture usually cannot be performed at home at midnight, and the patient must be hospitalized. Also, variable reference ranges have been suggested based on the sleeping state, length of hospital stay, and desired sensitivity and specificity (4).

Corticotropin Releasing Hormone (CRH) Stimulation Test: Most pituitary tumors will respond to CRH stimulation with increased ACTH and increased cortisol. 100 ug of synthetic ovine of human CRH is administered intravenously and ACTH and cortisol are measured over time (1). In most ACTH-dependent instances, a significant increase in both hormones will occur. In ACTH-independent cases of CS, ACTH and cortisol levels remain unchanged (4). Suggested diagnostic changes are an increase in baseline ACTH of 35-50% or an increase in cortisol of 14-20% above baseline; however, there is no consensus for the interpretation of this test (2). Ectopic ACTH-producing lesions may also respond to CRH stimulation. Besides CRH, other stimulating compounds being evaluated include desmopressin, GH secretagogues, and loperamide (2).

High Dose Dexamethasone Suppression Test: This evaluation can be performed with a two day course of 2mg every 6 hours, an 8 mg overnight dose, or a 4-7mg intravenous dose (2). High doses of glucocorticoid should partially suppress ACTH secretion from most corticotroph pituitary tumors while ectopic tumors resist feedback inhibition (1). In adrenal CS, ACTH is already suppressed and cortisol secretion will not diminish because it is already autonomous. No protocol variation, however, allows for complete discrimination between ectopic ACTH and CD (2).

Search for Occult Ectopic ACTH-secreting Tumors: Computed tomography or magnetic resonance imaging of the neck, thorax, and abdomen should be performed to identify possible ectopic sources of ACTH should IPSS fail to localize centrally. Selective venous catheterization with measurement of ACTH levels throughout the venous system may also be useful for localization (10). Most commonly, ectopic ACTH secretion stems from small cell carcinoma of the lung or bronchial carcinoid (4). Other neuroendocrine neoplasms like thymic or pancreatic carcinoids, islet cell tumors, medullary thyroid cancer, and pheochromocytomas are also possible sources (2). Patients with more indolent lesions may be indistinguishable from those with CD (4).


Therapy is dependent upon the form of CS or CD diagnosed. Ideally, treatment should normalize ACTH and cortisol secretion and reverse clinical symptoms while preserving pituitary function (2,10)

Cushing's Disease: If a circumscribed adenoma has been identified, localized transsphenoidal resection is recommended. For a diagrammatic representation of transsphenoidal surgery, see . If a circumscribed lesion has not been identified and future fertility is not an issue, subtotal resection of 85-90% of the anterior pituitary is indicated (12). Whenever possible, the procedure should be performed by a neurosurgeon experienced in the technique. Patients who are concerned about future fertility, in whom a tumor is not identified, or who remain symptomatic postoperatively may undergo pituitary irradiation, either conventional or stereotactic (6).

As a bridge to surgery, when a patient refuses surgery, or when symptoms persist postoperatively, medical management is aimed at decreasing adrenal steroid production. Ketoconazole is the most commonly used competitive steroid synthesis inhibitor. Other options include aminoglutethimide, metyrapone, etomidate, and trilostane. If pharmacologic blockade of cortisol synthesis is insufficient, ACTH secretion may override the inhibition and hypercortisolemia will persist (6,10).

Alternatively, when medical management fails and symptoms persist postoperatively, adrenalectomy should be considered. This can be accomplished medically with administration of mitotane, an isomer of the insecticide DDT, which induces adrenal cortical cell death and blocks the first step of cortisol synthesis. Also, in the case of adrenal carcinoma, mitotane can be used as adjunctive post-surgical therapy despite the lack of evidence that it prolongs survival significantly. Mitotane is a difficult drug to administer: it is given in multiple divided daily doses, and dosage must be carefully titrated to maximize effect and minimize side effects such as diarrhea, vomiting, lethargy, and somnolence. Bilateral surgical adrenalectomy offers rapid cure and a cure rate close to 100%. The procedure itself is not without morbidity, and patients who did not receive pituitary irradiation are later at risk for developing Nelson's syndrome: rapid pituitary corticotroph tumor progression (11).

Exogenous Cushing's syndrome: Exogenous glucocorticoid should be withdrawn slowly to allow time for recovery of hypothalamic-pituitary adrenal suppression (Up To Date:

Ectopic ACTH and CRH Syndromes: If surgical resection is not possible, hypercortisolism can be treated with medical therapy or bilateral adrenalectomy as described above (11).

Primary Adrenal Disease: Adrenal adenomas are reliably cured surgically. In the case of adrenal carcinoma, which frequently recurs and is unresponsive to chemo- and radiotherapy, mitotane may prevent recurrence in operable cases or provide palliative therapy in inoperable instances (11).

Criteria for Cure and Prognosis:

Low serum cortisol, less than 1.8 ug/dL, two weeks post-surgery is a good indicator of remission. Even then, late relapses can occur (2). Circumstances that may indicate relapse include non-suppressed post-operative serum cortisol or brief duration of post-surgical hypoadrenalism (10). Dynamic tests may suggest increased or decreased risk of recurrence, but cannot predict outcome in a single patient (2). One simple screening test of normal glucocorticoid secretion is the rapid ACTH stimulation test. At any time of the day, baseline plasma cortisol is measured, 0.25mg of cosyntropin (synthetic ACTH) is administered, and plasma cortisol levels are measured again at 30 and 60 minutes later. The most precise cut-off for normal response is a cortisol >18 ug/dL, and the minimal cut-off for stimulated normal is >7ug/dL above baseline (11).

Post-pituitary-surgical remission rates vary greatly from 42-86% with centers with greater caseloads having better outcomes (6). Nevertheless, the risk of relapse persists, with recurrence in approximately 25% of patients at 10 years (2).

Following pituitary radiation therapy, remission rates are good: 50-83% for conventional therapy and 63-83% for stereotactic methods like gamma knife radiosurgery. However, cortisol normalization can take up to two and a half years and medical therapy may be required in the interim. Ophthalmoplegia, cerebrovascular events, and second tumors (2.4% risk at 20 years) are rare complications of traditional radiotherapy (10).

Left untreated, the complications of CS may significantly alter life expectancy. Most affected patients develop at least some features of metabolic syndrome, including the increased cardiovascular risk, osteoporosis with risk of fracture, and psychiatric or mood effects. Therefore, it is particularly important that the appropriate diagnostic evaluation and clinical or surgical management be performed (2).


Given the significant rate of relapse even 10 years following curative surgical therapy, follow-up is mandatory. Regular evaluation of the hypothalamic-pituitary-adrenal axis with UFC, midnight cortisol, or suppression testing must be performed and confirmed by dynamic testing and pituitary imaging. Patients should also be monitored for development of additional pituitary hormone deficiencies. In order to monitor for Nelson's syndrome, plasma ACTH should be assessed annually. Also, appropriate medical therapy and monitoring for the aforementioned possible complications of CS should be performed (10).


Post-operatively, serum cortisol was down to below the normal range, and the patient continued on hydrocortisone replacement therapy. Several subsequent evaluations demonstrated normal residual pituitary function and persistent low-normal cortisol and high-normal ACTH. As of April 2008, the patient's weight had decreased to 217lbs with a BMI of 41. (Table 4)

Also on the 18th of April, the patient underwent an ACTH stimulation test to evaluate residual function of her hypothalamic-pituitary axis (Table 5). Following administration of ACTH, her serum cortisol only increased by 6ug/dL, just shy of the minimal cut-off for normal pituitary response. She will continue on hydrocortisone supplementation and have her ACTH and cortisol monitored every 6 weeks.


A very brief biography of Harvey Williams Cushing (1869-1939) can of course be found on Wikipedia. If you prefer a slightly more authoritative and less condensed source, John F. Fulton wrote the definitive Biography of Harvey Cushing (available online at Also, Michael Bliss penned Harvey Cushing: A Life in Surgery (Oxford Press 2005). A brief search on did turn up available copies of the Bliss biography and of Cushing's own The Pituitary Body and its Disorders (1910) and The Life of Sir William Osler (1925).


  1. Pivonello, R, et al. Cushing's Syndrome. Endocrinol Metab Clin N Am 2008; 37:135-149.
  2. Arnaldi, G, et al. Diagnosis and Complications of Cushing's Syndrome: A Consensus Statement. J Clin Endocrinol Metab 2003. 88:5593-5602.
  3. Lad, SP, et al. The role of inferior petrosal sinus sampling in the diagnostic localization of Cushing's disease. Neurosurg Focus 2007. 23(3):E2.
  4. Riemondo, G, et al. Laboratory differentiation of Cushing's syndrome. Clinica Chimica Acta 2008. 288:5-14.
  5. Becker, KL, ed. "Cushing Syndrome." Principles and Practice of Endocrinology and Metabolism. 3rd ed. Philadelphia: Lippincott Williams & Wilkins, 2001.
  6. Findling, JW, Raff, H. Clinical Review: Cushing's Syndrome: Important Issues in Diagnosis and Management. J Clin Endocrinol Metab 2006. 91(10):3746-3753.
  7. Rumboldt, Z. Pituitary Adenomas. Top Magn Reson Imaging. July 2005. 16(4):277-288.
  8. Miller, DL & Doppman, JL. Petrosal Sinus Sampling: Technique and Rationale. Radiology 1991. 178:37-47.
  9. Colao, A et al. Inferior petrosal sinus sampling in the differential diagnosis of Cushing's Syndrome: results of an Italian multicenter study. European Journal of Endocrinology 2001. 144: 499-507.
  10. De Martin, M et al. Cushing's disease. Pituitary 2006. 9:279-287.
  11. Williams, GH and Dluhy, RG. "Disorders of the Adrenal Cortex." Harrison's Principles of Internal Medicine. Eds E. Braunwald et al. 15th ed. New York: McGraw-Hill, 2001. 2084-2104.
  12. Testa, RM et al. The usefulness of combined biochemical tests in the diagnosis of Cushing's disease with negative pituitary magnetic resonance imaging. European Journal of Endocrinology 2007. 156:241-248.
  13. Swearingen, B et al. Diagnostic Errors after Inferior Petrosal Sinus Sampling. Journal of Clinical Endocrinology and Metabolism 2004. 89(8):3752-3763.

Contributed by Hannah Kastenbaum, MD with Mohamed A. Virji, MD, PhD

Case IndexCME Case StudiesFeedbackHome