Hypercortisolism is a disorder characterized by excessive circulating cortisol from either endogenous or exogenous sources, which can cause detrimental changes to nearly every organ system. Collectively, the ravages of hypercortisolism are best classified as Cushing’s syndrome (CS). The term Cushing’s syndrome is used for all causes of hypercortisolism, and the term Cushing’s disease is reserved specifically for cases of pituitary-dependent CS. The most common cause of CS is exogenous administration of glucocorticoids, which are used for their beneficial antiinflammatory, antineoplastic, and immunosuppressive effects. This chapter focuses solely on endogenous CS and is a summary of our recent review at the Mayo Clinic in Rochester, Minnesota.1

The incidence of CS is approximately two cases per 1 million persons per year.1 Patients who are not cured by surgical intervention or with malignant disease have a poor prognosis, with a standard mortality ratio of 3.8 to 5.0 compared with normal control subjects. Adrenocorticotropic hormone (ACTH)–dependent causes account for 80% of all cases of endogenous CS, with the majority of them (85%) attributed to pituitary-dependent causes; only 15% of cases are caused by ectopic production of ACTH or corticotropin-releasing hormone (CRH). The remaining 20% of cases of endogenous CS are caused by ACTH-independent CS. Specifically, 10% of all cases of endogenous CS are attributed to adrenal adenomas, 5% to 8% are attributed to adrenocortical carcinomas (ACCs), and the remaining 2% to 5% to ACTH-independent macronodular adrenal hyperplasia (AIMAH) and primary pigmented nodular adrenal disease (PPNAD) (Figure 14-1).1–3

Because of the multiplicity of causes, there is no single risk factor or set of risk factors associated with CS. However, a common factor that is present in individuals with CS is a female predilection, with the exception being CS caused by ectopic ACTH production, which is slightly more common in men. In our recent 10-year review of 288 patients who underwent surgery for CS at the Mayo Clinic, women accounted for 79% of all cases of pituitary-dependent CS, 81% of adrenocortical adenomas, and 90% of ACCs. However, among patients with ectopic ACTH syndrome, men are affected slightly more often than women at 54%.1

In 1912, Harvey Cushing described a 23-year-old woman, “Minnie G.,” with moon facies, obesity, amenorrhea, and hirsutism.4 It was not until 20 years later that he postulated that this “polyglandular syndrome” was caused by a primary pituitary abnormality that resulted in adrenal hyperplasia. This full-blown clinical picture is not always present, so one must maintain a high index of suspicion when this condition is suspected. In adults, weight gain and centripetal obesity are the most common signs. In children, short stature and generalized obesity should raise the suspicion of CS.2 Excessive fat deposition occurs preferentially over the thoracocervical spine, supraclavicular fossae, cheeks, and temporal regions, giving rise to descriptions of a “buffalo hump,” “moon-like facies,” or the visual effect of “disappearing ears” when the patient is viewed in a series of facial photographs. Patients are often troubled by menstrual irregularity, infertility, and loss of libido.2 Acne and hirsutism are common, particularly in women, because of excessive ACTH stimulation of the androgen-producing zona reticularis. The excess cortisol also has an inhibitory effect on gonadotropin production, resulting in hypogonadism.

More than half of patients with CS have psychiatric disturbances. The most common are agitated depression, lethargy, irritability, insomnia, and cognitive dysfunction. In extreme cases, severe paranoia and psychosis may be present.

In children, short stature is related to impaired linear growth in contrast to the loss of height seen in adults caused by vertebral compression fractures. Accelerated bone loss, osteoporosis, and skeletal fragility may result in asymptomatic rib and other atraumatic fractures. Aseptic necrosis of the femoral and humeral heads may also be seen. Proximal myopathy of the lower extremities and shoulder girdle are common and are a major discriminatory feature of CS. An inability to rise from a crouching position is often revealing.1,2

In the skin, cortisol excess results in thinning, separation, and exposure of underlying subcutaneous vascular tissue, giving rise to the plethoric features. Minimal trauma leads to “easy bruising,” which is exacerbated by delays in wound healing. The typical 1- to 2-cm, wide, red-purple striae are virtually pathognomonic. They are most frequently found on the abdomen but may also be seen on the upper thighs, breasts, and arms.1,2 Increased skin pigmentation, or “bronzing,” is common in individuals with ectopic ACTH syndrome because of stimulation of skin melanocyte receptors from high levels of circulating ACTH. Fungal nail and skin infections, systemic fungal infections, and other opportunistic infections may also be seen in those with glucocorticoid excess.1

Three quarters of patients with CS have hypertension as part of their severe metabolic syndrome. The well-recognized consequences of diabetes and dyslipidemia contribute to increased cardiovascular morbidity and mortality.1,2 Thromboembolic events and intestinal perforations are also more common in people with CS. Glucose intolerance and overt diabetes are seen in up to one third of patients with CS. Hypokalemic alkalosis, seen in 10% to 15% of patients with pituitary CS, is present in more than 95% of those with ectopic ACTH syndrome because of marked increases in cortisol levels and its effect on the mineralocorticoid receptor. In addition to suppressing pituitary gonadotropins, glucocorticoids directly inhibit Leydig cell function. Excess cortisol may increase fat deposition in the retroorbital regions, leading to exophthalmos, and in the epidural space, leading to neurologic sequelae from spinal cord or cauda equina compression. Excess cortisol may also increase intraocular pressure, resulting in glaucoma.1,2

The often insidious onset of CS is one of the most challenging problems in endocrinology. It is common for the signs and symptoms to have been present for 3 to 5 years. In differentiating cases of CS from normal aging and simple weight gain, old photographs are particularly helpful. In obese individuals with type 2 diabetes who have poor glycemic control, the prevalence of CS may be as high as 4%.2 After CS has been recognized, appropriate and cost-effective laboratory testing are required for confirmation, subtype determination, and (when possible) definitive cure.

When hypercortisolism is suspected, our first test of choice is usually measurement of urinary-free cortisol (UFC) in a 24-hour collection. We also quantify urine creatinine for complete collection quality control. The suspicion for CS is elevated when the 24-hour UFC is greater than twice the upper limit of normal (normal range, 3.5 to 45 μg/24 hr or 9.7 to 124 nmol/d). Loss of diurnal variations in serum cortisol can be perturbed under many circumstances, including stress and hospitalization, and are thus less reliable than 24-hour urine collections. Even in patients with known CS, 10% to 15% of patients have one of four serial 24-hour UFCs in the normal range. All causes of CS can produce cortisol in an episodic fashion. Therefore, if suspicion is high based on the clinical assessment, repeated collections are indicated. Use of tandem mass spectrometry has avoided the falsely elevated 24-hour UFC caused by medications such as carbamazepine. Other scenarios in which the 24-hour UFC may be falsely elevated include high urine volumes (4 L/24 hr), severe illness, alcoholism, depression, and obstructive sleep apnea (OSA). When the clinical picture is consistent with CS and the baseline 24-hour UFC exceeds five times the upper limit of normal (300 μg or 828 nmol/d), no additional studies are needed to confirm CS (Figure 14-2).

An overnight 1-mg dexamethasone suppression test (DST) is a useful test for demonstrating the autonomy of an existing adrenocortical nodule.5 The 8 AM plasma cortisol level in normal patients will suppress to below 5 μg/dL (138 nmol/L) with this test. There are, however, many causes of cortisol nonsuppression with the overnight 1-mg DST, including patient error, increased corticosteroid-binding globulin (CBG) caused by estrogen therapy or pregnancy, alcoholism, OSA, depression, panic attacks, obsessive–compulsive disorder, obesity, drugs that accelerate dexamethasone metabolism (anticonvulsants, primidone, rifampin), renal failure, stress, cortisol assay insensitivity, and laboratory error. Another option is the 8-mg overnight DST. Normal patients should have a morning cortisol level of nearly 0. Failure to suppress is a positive test finding and warrants a 2-day, low-dose DST.

With equivocal findings and a 24-hour UFC below 300 μg (828 nmol/d), hypercortisolism should be confirmed with a 2-day, low-dose DST. Dexamethasone, 0.5 mg orally, is given every 6 hours for 48 hours. A 24-hour UFC collected during the second day of suppression of 3.5 μg (9.7 nmol/d) or above is consistent with the diagnosis of CS. This test is far from perfect, with a sensitivity of 79%, specificity of 74%, and overall accuracy of 71%. The test is most efficacious for ruling out CS in patients in whom the index of suspicion for CS is low. In addition, some patients with mild pituitary-dependent CS may suppress with a low-dose DST.5

In an effort to correct the suppression with a low-dose DST observed in some patients with pituitary CS, the CRH DST test was developed. A serum cortisol concentration above 1.4 μg/dL (37.6 nmol/L) at 15 minutes after CRH administration is consistent with CS. Although purported to be highly accurate, false-positive results have been seen in our clinic.

Other diagnostic tests include the measurement of late night plasma or salivary cortisol. A midnight sleeping plasma cortisol concentration above 1.8 μg/dL (50 nmol/L) is 100% sensitive in patients with CS.1 However, for the test to be precise, hospitalization is usually required. Alternatively, measurement of salivary cortisol correlates well with serum cortisol levels. This is performed by giving the patient instructions to chew on a special cotton swab for 2 minutes and then place it in a specially designed container. Salivary cortisol levels obtained at 11 PM that are above 100 ng/dL (2.6 nmol/L) are highly sensitive for CS. This test is becoming increasingly more popular and is used as a first-line screening test in some centers despite problems with reproducibility, sensitivity, and precision of the assay.

No single test is satisfactory for case finding and confirmation of CS. The clinician should be certain of the diagnosis of CS before proceeding to subtype evaluation and localization. Patients whose clinical assessment suggests a possible diagnosis of CS when exogenous sources have been excluded should be evaluated with multiple tests for diagnostic confirmation. These include, at a minimum, a 24-hour UFC aided by an 11 PM salivary cortisol to be followed by the two-day, low-dose DST or CRH DST test. Patients with clinically evident CS and a 24-hour UFC above 300 μg (828 nmol/d) need no further confirmatory studies. Patients with an incidentally discovered adrenal mass and an 8 AM plasma cortisol concentration below 5 μg/dL (138 nmol/L) after an overnight 1-mg DST also need no further testing because autonomous cortisol production has been ruled out.

Subtype Evaluation and Imaging