This patient provides an interesting example of the diagnostic evaluation for a patient with Cushing's syndrome. Cushing's syndrome is an endocrine disorder caused by hypercortisolism that was originally described by Dr. Harvey Cushing in 1932 (1). The syndrome can be caused by the exogenous administration of glucocorticoids or endogenously by an abnormality in the hypothalamic-pituitary-adrenal (HPA) axis. The later can be subdivided into ACTH dependent and independent types. The most common ACTH dependent cause is increased ACTH secretion from the pituitary, often due to a pituitary adenoma (2). The second most common ACTH dependent cause is ectopic production of ACTH from tumors including small-cell carcinomas of the lung, bronchial carcinoids, and pancreatic endocrine neoplasms (3). Additionally, rare cases of ectopic CRH production have been reported (4). Common ACTH independent causes of endogenous Cushing's syndrome include adrenal adenoma and adrenal carcinoma (2).
The patient initially presented with many textbook features of Cushing's syndrome including central adiposity, Cushingoid facies, a large dorsal fat pad, hirsutism, striae, and proximal muscle weakness. Additionally, she had hyperglycemia and hypokalemia. According to the 2008 Endocrine Society Clinical Guidelines, the initial testing for a patient suspected of having Cushing's syndrome (after excluding exogenous glucocorticoid use) is a urinary free cortisol, late-night salivary cortisol, a 1mg overnight dexamethasone suppression test, or a longer low-dose dexamethasone suppression test (5). In this case, the patient had elevated urinary free cortisol (greater than 1000 µg/day) and a low dose dexamethasone suppression test showed failure of cortisol suppression (AM cortisol 56 µg/dL).
Once the diagnosis of Cushing's syndrome is established, the next step is to evaluate whether the cortisol production is ACTH dependent. An ACTH dependent cause will have an ACTH level greater than 10 pg/mL, while an independent cause will have a level less than 5 pg/mL (6). In the case of this patient, ACTH levels were very high (161 pg/mL) indicating an ACTH dependent cause.
Next, the endocrinology team evaluating this patient decided to perform a dedicated pituitary MRI because they felt a pituitary etiology to be most likely in a young female. Unfortunately, the MRI was equivocal, revealing a possible adenoma within the pituitary.
In order to help differentiate a pituitary from an ectopic source of ACTH, an overnight high-dose dexamethasone suppression test was performed. This test involves administering an 8mg dose of dexamethasone in the evening and then measuring the cortisol level in the morning. A decrease in cortisol of 50% or more is indicative of a pituitary source. It should be noted that 20-30% of patients with a pituitary source will have less than 50% cortisol suppression and 20-30% of patients with an ectopic source will have greater than 50% suppression (7). In the case of this patient, there was no suppression (AM cortisol 116 µg/dL), indicating an ectopic source of ACTH production.
At this point, the patient experienced gastrointestinal bleeding and endoscopy revealed numerous duodenal ulcers. This raised suspicion for Zollinger-Ellison syndrome, so serum gastrin and chromagranin were assayed and found to be extremely elevated at 1599 pg/mL and 142 ng/mL, respectively. Subsequently, an endoscopic ultrasound revealed a pancreatic mass, which was biopsied and found to be a neuroendocrine neoplasm. After this discovery, it was suspected that the liver lesion was a metastasis of this neoplasm.
While the workup of the pancreatic lesion was proceeding, a CRH stimulation test was performed. This test can also be used to help differentiate a pituitary from an ectopic cause of ACTH production. The basis of the test is that pituitary sources of ACTH generally respond to CRH with an increase in cortisol, whereas ectopic sources do not (8). There is no consensus on the interpretation of this test, as some have used a post CRH-stimulation ACTH increase of greater than 35% and a cortisol increase greater than 20% to diagnose Cushing's disease, whereas other groups have found an ACTH increase greater than 50% to have the greatest diagnostic accuracy (9-10). In this case, the patient responded to CRH with an ACTH increase of 37% and a cortisol increase of 51%. These results were borderline, but suggestive of a pituitary etiology.
Given the failure of the high-dose of dexamethasone to suppress cortisol and the discovery of the pancreatic endocrine neoplasm, the endocrinology team felt that the Cushing's syndrome was caused by ectopic production of hormone; however, it was necessary to rule out a pituitary etiology because of the possible pituitary lesion discovered on the MRI and the CRH stimulation test that was suggestive of a pituitary cause. The decision was then made for the patient to undergo inferior petrosal sinus sampling (IPSS), a procedure in which catheters are inserted peripherally through the femoral veins and advanced to the inferior petrosal sinus (IPS) on each side. Patients with an ACTH producing pituitary lesion will have higher levels of ACTH in the IPS compared to the peripheral blood because it receives drainage directly from the pituitary. Measurements are often taken at baseline and after the stimulation of ACTH release with CRH. In a study involving 215 patients, an IPS to peripheral ratio of greater than 2 in baseline samples identified patients with Cushing's disease with 95% sensitivity and 100% specificity and a post CRH IPS to peripheral ratio of greater than 3 identified patients with Cushing's disease with a 100% sensitivity and 100% specificity (11). In this case, the ratio of the baseline IPS to peripheral blood ACTH level was less than 2 and the post CRH stimulation ratio was less than 3. These findings substantiate that her Cushing's syndrome was due to ectopic ACTH production from the pancreatic endocrine tumor and not from a pituitary lesion.
The patient's elevated cortisol was treated with ketoconazole, which inhibits multiple enzymes in the cortisol synthesis pathway. The patient's cortisol levels normalized; however, she developed elevated liver enzymes from the ketoconazole and was switched to metyrapone which also inhibits cortisol synthesis. Unfortunately her hospital course was complicated by a large saddle pulmonary embolism. Of note, she was later discovered to be heterozygous for the Factor V Leiden mutation. Resection of the lesions was delayed due to the patient's high surgical risk secondary to the pulmonary embolism, so the patient underwent chemoembolization of the liver lesion.
Interestingly, this patient was found to have multiple endocrine neoplasia type I after DNA sequence analysis revealed a pathogenic mutation in MEN1. This syndrome is associated with pancreatic endocrine neoplasms as well as tumors of the pituitary and parathyroid. There was no history of similar tumors in this patient's family.
Contributed by Stephen Hastings, MD and Mohamed A. Virji, MD, PhD