David A. Goldfarb
Contemporary evaluation and management of Cushing’s syndrome
Abstract Cushing’s syndrome, characterized by unreg- ulated cortisol secretion, may be caused by a variety of adrenal, pituitary, or other tumors. The best biochemi- cal test for establishing the diagnosis is determination of 24-h urinary free cortisol. The specific causes for Cush- ing’s syndrome may be further differentiated by plasma adrenocorticotrophic hormone (ACTH). Primary adre- nal cortical diseases are associated with low levels of ACTH and are considered ACTH-independent. Pitui- tary disease and the ectopic ACTH syndrome are asso- ciated with normal or elevated ACTH levels and are considered ACTH-dependent. Adrenal forms of Cush- ing’s syndrome may result from either adenoma or carcinoma. The diagnostic approach to Cushing’s syn- drome and the clinical, biochemical, and radiographic features that distiguish adrenal adenoma and carcinoma are the subjects of this paper.
65-75% of CS [14, 15]. Most cases of Cushing’s disease are the result of pituitary adenomas; however, cortico- trope hyperplasia is responsible for a small minority of cases. Ectopic production of ACTH from a variety of tumors (bronchial carciniod, thymoma, oat-cell carci- noma, pheochromocytoma, islet-cell tumor, and pros- tate cancer) accounts for 10-15% of CS. Primary adrenocortical diseases account for the remaining 20- 30% of CS, including benign adenoma (10-15%), ad- renocortical carcinoma (5-10%), and adenomatous hy- perplasia (5%).
The purpose of this review is to present a contem- porary approach to the evaluation and management of patients with Cushing’s syndrome, emphasizing the primary adrenal etiologies pertinent to urologists.
Harvey Cushing initially described the constellation of features, including obesity, diabetes, and adrenal hy- perplasia, and then observed basophilic pituitary ade- nomata in 75% of these patients [2]. Cushing’s syndrome (CS) is characterized by excessive cortisol se- cretion that has escaped the normal regulatory mecha- nisms. It is now recognized in association with a wide variety of primary pathologic entities. Nonetheless, etiologies may be broadly classified into adrenocortico- trophic hormone (ACTH)-dependent or ACTH-inde- pendent pathologies. Of the ACTH-dependent causes, Cushing’s disease (excessive pituitary production of ACTH) is the most common etiology, accounting for
Diagnosis
Clinically, patients may present with a variety of signs and symptoms, including round facies, proximal muscle weakness, obesity (mainly truncal), thin skin, easy bruisability, purple stria, psychiatric symptoms, hyper- tension, and diabetes. Although no single sign or symptom is sensitive or specific enough for the diagno- sis, when a patient presents with the combination of proximal muscle weakness, thin skin, and any of the other signs or symptoms noted above a biochemical evaluation should be pursued [13].
These tests are directed at identifying excessive cor- tisol secretion or abnormal regulation of the hypotha- lamic-pituitary-adrenal (HPA) axis. The normal diurnal variation in plasma cortisol is lost in CS, rendering the use of random plasma cortisol unreliable for diagnosis. Since cortisol release is episodic, an integrated repre- sentation of plasma cortisol over a prolonged period yields the best estimation of secretion. Determination of 24-h urinary free cortisol is the best indicator of inte- grated cortisol secretion and is the most commonly used laboratory test to document excessive cortisol secretion
D. A. Goldfarb Cleveland Clinic Foundation, Department of Urology (A110), 9500 Euclid Avenue, Cleveland, OH 44195, USA e-mail: goldfad@cesmtp.ccf.org, Tel .: + 1-216-444-8726, Fax: +1-216-444-9375
[3, 4, 18]. Several samples should be collected, with si- multaneous creatinine measurements being done to in- sure an accurate collection. False-positive results can occur in individuals with alcoholism, depression, or se- vere illness; but patients with these conditions usually lack the somatic features of CS.
Dexamethasone suppression tests (DST) represent another method of screening for CS. The normal re- sponse to exogenous dexamethasone is suppression of cortisol secretion. In CS, abnormal regulation of the HPA axis is associated with resistance to the effects of exogenous steroids. As a result, patients with CS do not suppress cortisol production after administration of dexamethasone. The simplest DST (overnight DST) in- volves the measurement of plasma cortisol at 8:00 a.m. following the oral administration of a 1-mg dose of dexamethasone at 11:00 on the previous night. The 2-day low-dose DST (0.5 mg every 6 h) has been advo- cated as a more accurate test but is logistically more cumbersome to perform on an outpatient basis. Failure to suppress plasma and/or urinary cortisol after the overnight or low-dose DST indicates CS. Due to the episodic secretion and spontaneous remissions that are observed in CS, if the diagnosis is strongly suspected and initial testing has been nondiagnostic, repeat testing may be required to confirm the diagnosis.
Differential diagnosis
Once the diagnosis of CS has been confirmed by the screening tests, the specific pathology needs to be iden- tified for the development of an appropriate treatment plan (see Fig. 1). Several plasma ACTH levels should be obtained [4, 15]. A low or undetectable level of ACTH suggests primary adrenocortical disease, and in such cases, computed tomography (CT) or magnetic reso- nance imaging (MRI) of the adrenals should be ob- tained. These studies reliably demonstrate the presence of a unilateral adenoma or carcinoma. MRI has the
added advantage of discriminating between adenoma and carcinoma on the basis of the signal intensity of T2- weighted images [12]. Adenomas are isointense with the liver on T2-weighted images, whereas carcinomas dem- onstrate an intermediate, heterogeneous increase in sig- nal intensity. In the setting of biochemically established ACTH-independent CS the size of the adrenal mass is the most important feature distinguishing adenoma from carcinoma [3]; hence, in most cases a CT scan is all that is required.
When the ACTH level is normal or mildly elevated, pituitary disease is the most likely diagnosis. MRI of the pituitary should be performed; however, it has a sensi- tivity of only 71% for the detection of an adenoma [11]. Several biochemical tests can help differentiate the var- ious causes of ACTH-dependent CS. A high-dose DST (2.0 mg every 6 h) may be used in this setting [4, 15]. In Cushing’s disease there is only a relative degree of re- sistance to the effects of exogenous steroids; thus most of these patients show suppression of cortisol secretion with the high-dose DST. In contrast, patients with the ectopic ACTH syndrome usually demonstrate no sup- pression with either the low-dose or the high-dose DST. The high-dose DST has been associated with a false- positive and false-negative rate of approximately 10%. Another biochemical assessment that has recently been described is the corticotrophin-releasing hormone (CRH) test [4, 15]. Pituitary adenomas express receptors for CRH, and patients demonstrate an increase in ACTH in response to CRH. Tumors responsible for the ectopic ACTH syndrome usually do not express recep- tors for CRH and do not demonstrate increases in ACTH in response to CRH administration. In cases that remain equivocal the most accurate test to differentiate between pituitary disease and other forms of CS is bi- lateral petrosal venous sinus sampling for ACTH before and after stimulation with CRH [9]. This, however, is an invasive examination and is accompanied by risks sim- ilar to those involved with cereberal angiography. A petrosal vein/peripheral blood ACTH ratio of greater
24-Hour Urinary Free Cortisol / Low Dose DST Plasma ACTH X 2
Low
Normal - Elevated
≥ 2X Normal
CT Adrenal MRI Adrenal
MRI Pituitary High Dose DST Petrosal Sinus Sampling
Adrenal Adenoma Adrenal Carcinoma
Pituitary Adenoma Pituitary Hyperplasia
Ectopic ACTH producing tumor
than 2.0 suggests the presence of pituitary disease. In addition to confirming a diagnosis, this test may also localize pituitary microadenoma.
When the ACTH value is > 2 times normal the ec- topic ACTH syndrome is likely; however, radiographic evaluation for a pituitary tumor should be performed. A high-dose DST or CRH test may be necessary. In equivocal cases, petrosal sinus venous sampling may also be required. When the biochemistry data suggest the ectopic ACTH syndrome a CT scan of the abdomen and chest should be performed as the initial radio- graphic evaluation for the variety of tumors responsible for this syndrome.
Adrenal tumors
The diagnosis of adrenal tumors is usually straightfor- ward. Plasma ACTH levels are well suppressed and an adrenal mass is identified on CT or MRI. The distinction between a benign adenoma and a malignant adrenal cortical neoplasm is important due to the difference in long-term patient survival. Ultimately, the only method certain to establish the difference is pathologic analysis [17]. Nonetheless, a reasonable assessment of the ma- lignant potential can be made using clinical, biochemi- cal, and radiographic information. A recent review of 40 patients (27 adenomas, 13 carcinomas) with ACTH-in- dependent CS was undertaken at our institution with the primary aim of establishing criteria for the preoperative delineation of adrenal adenomas from carcinomas [3].
The two demographic features that distinguished adenomas from carcinomas were age and tumor size. As compared with carcinoma patients, adenoma patients were significantly younger (39.6 versus 51.5 years) and their tumors were significantly smaller (3.3 versus 8.6 cm). Women constituted the majority of patients in both groups, and left-sided tumors accounted for 70% of cases in both groups. There was no specific physical finding that differentiated adenomas from carcinomas.
All patients had elevated 24-h urinary free cortisol values and low/normal ACTH levels. There was no specific biochemical abnormality that would identify either adenoma or carcinoma patients. Nevertheless, urinary free cortisol, 17-ketosteroid (17-KS), de- hydroepiandrosterone sulfate (DHEAS-S), and lactate dehydrogenase (LDH) levels tended to be higher in carcinoma patients. A pure biochemical syndrome of glucocorticoid excess without elevation of 17-KS, DHEA-S, testosterone, or aldosterone was present in 68% of adenoma patients as opposed to only 8% of carcinoma patients.
Some groups have noted the presence of virilization/ hirsutism as an important clinical clue suggesting the presence of carcinoma [10]. This is based on the premise that carcinomas are less effective than adenomas at converting steroid precursors to glucocorticoids. In our recent review this was not a discriminating feature [3]. Whereas most carcinoma patients presented with a
mixed endocrine syndrome (92%), adenoma patients commonly had virilization (93%) and 32% presented with a mixed endocrine syndrome as determined by biochemical analysis. There were a few noteworthy ob- servations. Of the adenoma patients in whom DHEA-S was measured, 44% had subnormal values, and in the two patients with elevated levels the values were only mildly raised. In contrast, only one carcinoma patient had a subnormal level of DHEA-S, and 50% of the carcinoma patients had elevations that exceeded 4 times the normal value. Overall, the magnitude of the bio- chemical abnormalities were more severe for the carci- noma patients. These findings support the notion that adenomas in general are more efficient than carcinomas at converting steroid precursors to glucocorticoids; however, there remains significant tumor-to-tumor variability. Although steadfast rules do not exist with regard to the preoperative hormonal profile capable of predicting the presence of adenoma versus that of car- cinoma, it is likely that patients with ACTH-indepen- dent CS, normal 17-KS levels, and subnormal DHEA-S values have an adenoma, whereas those with greatly elevated 17-KS and DHEA-S levels have a carcinoma. Although steroid profiling may not dramatically en- hance the management of patients with a small adrenal mass that is likely to be an adenoma, it is important for patients with suspected carcinoma because the steroid values can be helpful in the follow-up of these patients, serving as markers of recurrent disease.
The radiographic features of an adrenal tumor are very important in the assessment of its malignant po- tential. In patients with ACTH-independent CS (as well as other adrenal tumors), tumor size is the most im- portant feature for differentiating adenomas and carci- nomas [3]. Tumors measuring >6 cm have a significant risk for malignancy. In our recent study, tumor size was one of the only features capable of distinguishing ade- nomas from carcinomas. The CT findings of adenoma typically show a small mass (<6 cm) in association with a low degree of attenuation and little, if any, calcifica- tion. On the other hand, carcinomas are typically larger (>6 cm) and inhomogeneous, demonstrate central ne- crosis, and may contain calcifications. MRI is also a good method for assessing the malignant potential of an adrenal mass [12]. Adenomas are isointense as compared with the liver on T2-weighted images, whereas carcino- mas are hyperintense and inhomogeneous. Nonetheless, there is as much as 30% overlap in these MRI charac- teristics between adenomas and carcinomas. In the set- ting of biochemically established ACTH-independent CS, CT is usually all that is required because a tumor is likely to be removed surgically, regardless of its size.
The presence of a hyperplastic adrenal gland on CT suggests an ACTH effect. It is important that this be recognized so that inappropriate adrenalectomy is not undertaken. Nonetheless, there is an infrequent cause of ACTH-independent CS that deserves mention. In mas- sive macronodular hyperplasia the normal adrenal is completely replaced by large nodules [4]. The etiology of
this unusual pathology is not clear. Bilateral adrenalec- tomy is the appropriate treatment as this disorder rep- resents a form of autonomous cortisol secretion that is independent of ACTH effect.
Surgery for adrenal tumors has evolved in the past 10 years. The most important surgical advance during this era has been laparoscopic adrenalectomy [5]. For large adrenal tumors where the primary diagnostic consider- ation is carcinoma an anterior transabdominal approach is useful [7]. For particularly large tumors a thoracoabdominal approach may be required. Using these approaches, complete surgical staging may be achieved and any required adjunctive surgery may also be addressed (i.e., nephrectomy, bowel resection, liver re- section, vena caval thrombectomy). For smaller tumors that are likely to be adenoma, laparoscopic adrenalec- tomy is rapidly becoming the surgical technique of choice, although an extraperitoneal (flank or posterior) approach [8, 16] may be used by surgeons unfamiliar with advanced laparoscopy techniques. In skilled hands, laparoscopic adrenalectomy may be performed with di- minished postoperative pain and a shorter convalescence period [5]. Refinements in laparoscopy techniques now permit a needlescopic approach [6]. This is associated with improvements in operative time, blood loss, con- valescence, narcotic use, and cosmesis as compared with conventional laparoscopic adrenalectomy. When a nee- dlescopic approach is used, >90% of patients are dis- charged from the hospital on the 1st postoperative day. Unless there is a specific contraindication, all patients with ACTH-independent CS and a small (<6 cm) adre- nal mass should be offered laparoscopic adrenalectomy.
In our experience with 27 adenoma patients there has been no post-operative recurrence [3]. One patient died of hypoglycemia, emphasizing the need for careful post- operative metabolic surveillance in these patients. Since the contralateral adrenal is suppressed in ACTH-inde- pendent CS, all patients require perioperative steroid replacement. The average duration of steroid depen- dence was 16.8 months; however, 25% (7/27) of patients never became steroid-free. This has been observed by other investigators and is an important part of the pre- operative counseling of patients.
The prognosis for carcinoma patients is not favorable, with the 3- and 5-year survival being 41.5% and 31.2%, respectively. The adrenolytic agent mitotane is useful for the control of symptomatic CS from unresectable or metastatic adrenal cortical carcinoma. Patients receiving mitotane need to remain on replacement steroids. The role of mitotane as adjuvant treatment in completely resected adrenal cortical carcinoma is controversial as a survival advantage has not been demonstrated [1].
Pituitary disease/ectopic ACTH syndrome
The primary mode of treatment for pituitary adenomas is transsphenoidal surgery [15]. This will result in the cure of CS in 80-95% of patients. Radiation therapy is
an option for patients who fail to respond to surgery. Adrenolytic medications or bilateral adrenalectomy would be indicated only for individuals who have failed both surgery and radiation. Bilateral adrenalectomy in the setting of Cushing’s disease has been associated with Nelson’s syndrome. This is characterized by elevated plasma ACTH levels, skin pigmentation, and the de- velopment of an invasive pituitary tumor.
Management of the ectopic ACTH syndrome in- volves identification of the source of its production and treatment of the primary tumor. Sometimes these tu- mors are quite small and present diagnostic dilemmas. For these cases, medical adrenalectomy using mitotane, aminoglutethimide, or ketoconazole may be required. Only in rare circumstances would bilateral adrenalec- tomy be required if a primary tumor could not be identified and medical forms of treatment had failed.
References
1. Bodie B, Novick AC, Pontes JE, et al (1989) The Cleveland Clinic experience with adrenal cortical carcinoma. J Urol 141: 257-260
2. Cushing H (1932) The basophil adenomas of the pituitary body and their clinical manifestations (pituitary basophilism). Bull Johns Hopkins Hosp 50: 137-195
3. Daitch JA, Goldfarb DA, Novick AC (1997) Cleveland Clinic experience with adrenal Cushing’s syndrome. J Urol 158: 2051- 2055
4. Findling JW, Doppman JL (1994) Biochemical and radiologi- cal diagnosis of Cushing’s syndrome. Endocrinol Metab Clin North Am 23: 511-537
5. Gill I, Clayman R, McDougall E (1995) Advances in urologic laproscopic surgery. J Urol 153: 666-679
6. Gill IS, Soble JJ, Sung GT, et al (1998) Needlescopic adre- nalectomy: comparison with conventional laparoscopic adre- nalectomy. Urology (in press)
7. Guz BV, Straffon RA, Novick AC (1989) Operative approaches to the adrenal gland, Urol Clin North Am 16: 527-534
8. Novick AC, Straffon RA, Kaylor W, et al (1989) Posterior transthoracic approach for adrenal surgery. J Urol 141: 254-256
9. Oldfield EH, Doppman JL, Nieman LK, et al (1991) Petrosal sinus sampling with and without corticotropin-releasing hor- mone for the differential diagnosis of Cushing’s syndrome. N Engl J Med 325: 897-905
10. Orth DN (1995) Cushing’s syndrome. N Engl J Med 332: 791
11. Peck WW, Dillon WP, Norman D, et al (1989) High-resolution MR imaging of pituitary microadenomas at 1.5T: experience with Cushing disease. Am J Radiol 152: 145-151
12. Reinig JW, Doppman JL, Dwyer AJ, et al (1986) MRI of in- determinate adrenal masses. AJR 147: 493-496
13. Sheeler LR (1988) Cushing’s syndrome - 1988. Cleve Clin J Med 55: 329-337
14. Sheeler LR (1989) Cushing’s syndrome. Urol Clin North Am 16: 447-455
15. Tsigos C, Chrousos GP (1996) Differential diagnosis and management of Cushing’s syndrome. Annu Rev Med 47: 443- 461
16. Vaughan ED, Phillips H (1987) Modified posterior approach for right adrenalectomy. Surg Gynecol Obstet 165: 453-455
17. Weiss LM (1984) Comparative histologic study of 43 metas- tasizing and nonmetastasizing adrenocortical tumors. Am J Surg Pathol 8: 163-169
18. Zadik Z, Lacerda L de, Camargo LA de, et al (1980) A com- parative study of urinary 17-hydroxycorticosteroids, urine free cortisol, and the integrated concentration of plasma cortisol. J Clin Endocrinol Metab 51: 1099-1101