0021 -972X/79/4805-0760$02.00/0 Jouriftl of Clinical Endocrinology and Metabolism Copyright © 1979 by The Endocrine Society

Vol. 48, No. 5 Printed in U.S.A.

Circadian Cortisol Secretory Rhythms in Cushing's Disease* ROBERT M. BOYAR, MICHAEL WITKIN, ALAN CARRUTH, AND JACK RAMSEY Departments of Internal Medicine and Pediatrics, University of Texas Health Science Center at Dallas, Dallas, Texas 75235; and the Ochsner Clinic (M. WJ, New Orleans, Louisiana 70121

ABSTRACT. Plasma cortisol was measured at 20-min intervals for 24 h in eight patients with Cushing's disease and ACTHsecreting pituitary tumors. The 24-h mean (±SD) cortisol level

control subjects; there was no significant difference in the number of cortisol secretory episodes in the patients us. control subjects. Three of the patients were restudied after successful

was 25.6 ± 11.3 /ug/dl (range, 15.5-40.6), which was significantly

resection of their ACTH-secreting pituitary tumors. Two showed

higher than the level of normal control subjects (P < 0.01). The 24-h mean ACTH level varied between 22-107 pg/ml, with a mean ±SD of 63.4 ± 27.2. The mean ACTH level was higher than that of the control subjects but the difference was not statistically significant. The 24-h cortisol secretory pattern was characterized by an absence of the normal circadian variation and a failure of the plasma cortisol level to fall to less than 2 jug/dl between 2300-0300 h. The coefficient of variation, an expression of the amplitude of cortisol secretory episodes, was significantly decreased in patients with Cushing's disease compared to normal

normalization of their 24-h circadian cortisol patterns and normal metyrapone responses. In the third, the 24-h mean cortisol level was normal, but the circadian cortisol rhythm remained abnormal. This patient had diminished ACTH reserve, demonstrated by a subnormal response to metyrapone. Additional studies will be required to determine if normalization of the circadian cortisol rhythm occurs in all patients with Cushing's disease who are cured after transsphenoidal microsurgery and who also show normal ACTH reserve. (J Clin Endocrinol Metab 48: 760, 1979)


of these patients will show recurrent microadenoma formation and Cushing's disease in the future if the primary defect is in the hypothalamus or higher central nervous system centers. One way to assess the possible presence of a primary extrapituitary abnormality is to measure the circadian variation of cortisol in patients cured after transsphenoidal adenomectomy at a time when normal ACTH-cortisol dynamics have been reestablished. Patients who show a normalization of their cortisol circadian rhythm and normal metyrapone responsiveness after transsphenoidal surgery presumably represent cases where the ACTH-secreting pituitary tumor may have arisen de novo, while those who show a persistently abnormal circadian rhythm can be considered examples of primary hypothalamic dysfunction producing secondary pituitary tumors. The present study was designed to define further the characteristics of the 24-h cortisol secretory patterns in patients with Cushing's disease due to pituitary tumors and to report circadian and stimulation studies in three of these patients who were cured after transsphenoidal adenomectomy.

USHING'S disease is associated with the inappropriate control of ACTH and cortisol production. Patients with this disorder have elevated cortisol levels (hypercortisolism) and normal to elevated ACTH levels (1). The underlying cause of this abnormal feedback regulation remains unknown. Several lines of evidence suggest that the disorder may be related to abnormal hypothalamic control mechanisms. Patients with Cushing's disease have been shown to have abnormal GH (2) and PRL secretion during sleep (3) as well as paradoxical ACTH responses to exogenous cortisol (4). Krieger et al. (5) have shown that some patients with Cushing's disease are successfully treated with cyproheptadine, presumably through an effect on central nervous system serotonergic mechanisms that normally restrain the secretion of corticotropin-releasing factor (CRF). Recently, Tyrrell and colleagues have shown that most patients with Cushing's disease have small ACTH-secreting microadenomas (6). Selective removal of these tumors results in amelioration of the hypercortisolism, but the long term effectiveness of this form of treatment has not yet been demonstrated. It is possible that some Received July 27, 1978. Address requests for reprints to: Dr. Michael Witkin, Ochsner Clinic, New Orleans, Louisiana 70121. * This work was supported in part by NIH Grants 1-KO4-HD-00153, l-ROl-HD-10909, and 5-MO1-RR-00633.

Materials and Methods Patients Seven patients with Cushing's disease were studied before transsphenoidal adenomectomy, and one subject (J.R.) was


The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 30 December 2014. at 09:40 For personal use only. No other uses without permission. . All rights reserved.

CORTISOL CIRCADIAN RHYTHMS studied after persistent Cushing's disease was documented after transsphenoidal hypophysectomy. Three patients were restudied after successful removal of pituitary microadenomas. The clinical data for these patients are given in Table 1. Four of the eight patients had abnormal polytomes of the sella turcica. All eight patients had elevated control 24-h urinary 17-hydroxycorticosteroids which failed to suppress to less than 3 mg/24 h after 2 days of low dose dexamethasone (2 mg/day) but did suppress to near 50% of control after 2 days of high dose dexamethasone (8 mg/day).


Cortisol secretory episodes were calculated utilizing the definition of Weitzman et al. (7), which states that a cortisol secretory episode has occurred when the plasma cortisol concentration rises at least 2 jug/dl and the next concentration is also higher than the initial one. The magnitude of cortisol secretory episodes in relation to the 24-h mean cortisol level was assessed by calculating the coefficient of variation (CV). This value was an expression of the degree to which the circadian variation and episodicity of cortisol secretion had been blunted. Results

Twenty-four-hour sampling studies

Clinical characteristics (Table 1)

The patients were admitted to the Clinical Research Center of the University of Texas Health Science Center at Dallas for study. Plasma samples were obtained at 20-min intervals for 24 h for measurement of plasma cortisol, as previously described (7-9). Patient L.G. was restudied 1 week and 8 months after removal of a basophilic pituitary microadenoma. Patients L.C. and C.S. were restudied 11 and 10 months, respectively, after removal of their pituitary microadenomas. These three patients also had metyrapone tests after cortisone replacement therapy had been discontinued. Plasma cortisol and ACTH Plasma cortisol was measured by RIA, utilizing the method and antibody of Abraham et al. (10). The 72 samples from each patient's 24-h study were assayed simultaneously to obviate interassay variation. The 24-h mean cortisol concentrations were obtained by averaging the 72 individual results from each patient's 24-h study. The intraassay and interassay coefficients of variation for the cortisol RIA were 6% and 8-10%, respectively. The sensitivity of the cortisol assay is 100 pg. The minimum detectable plasma value that is significantly different from 0 is 0.5 jug/dl utilizing 20 jul plasma. The cortisol values obtained during patient L.G.'s second study were measured on the linear portion of the standard curve between B/B o values of 0.56-0.75. The 24-h mean ACTH was measured on 24-h pooled samples obtained by combining an equal amount of plasma from each of the 72 samples. The ACTH was then measured by RIA (Nichols Institute, San Pedro, CA) (11). TABLE

All eight patients had clinical and laboratory evidence of Cushing's disease. Seven of the eight patients had identifiable pituitary tumors selectively removed, despite the fact that four of the eight patients had a radiographically normal sella turcica on polytomography. The eight patients underwent transsphenoidal pituitary exploration and resection, and in seven of the patients there was a decrease in the plasma cortisol level. One of the patients (J.R.) failed to show a clinical or laboratory remission despite an almost total hypophysectomy. She is presently receiving radiotherapy. Three of the patients are still receiving exogenous corticosteroids and have not been restudied. One of the patients (S.S.) expired during the postoperative period after pulmonary embolization. Twenty-four-hour cortisol secretory patterns (Fig. 1) The pretreatment 24-h cortisol secretory patterns were characterized by periods of episodic secretion occurring randomly throughout the 24-h period. There was no evidence of a circadian variation of cortisol secretory episodes and the plasma cortisol level did not fall to a value below 8.1 jug/dl at any time during the 24-h study in any of the patients. In six of the patients there was an elevated cortisol level in response to insertion of the catheter. Patient W.H. had a markedly elevated cortisol

1. Clinical data in patients with Cushing's disease Urinary steroids (mg/24 h) Sex

S.P. L.C. L.G. C.S. B.B. J.R. W.H. S.S. Normal


Age (yr)

35 57 34 24 41 32 38 64

Sella polytomes

Abnormal Normal Abnormal Normal Abnormal Normal Abnormal Normal





8mg 6




23.1 26.5 27.8 20.3 19.9 19.4 37.8 22.2 2-10

13.7 11.0 18.0

8.7 7.0 7.4 4.4 4.4 5.8

23.2 13.9 23.7 15.2 42.2 24.6 46.8

18.7 9.0

9.4 4.0





32.5 17.3 45.0

14.7 24.7





2 /xg/dl) between 2300-0300 h. This failure to normalize the cortisol circadian rhythm occurred despite the fact that the 24-h mean cortisol had returned to normal (5.5 jug/dl) and the clinical features of Cushing's syndrome had resolved. Because of persisting fatigue, the

The results of this study show that patients with Cushing's disease with proven ACTH-secreting pituitary tumors have 24-h circadian cortisol secretory patterns characterized by a loss of the normal circadian variation in cortisol secretory episodes and elevated cortisol values between 2300-0300 h. The 24-h mean plasma ACTH level is normal, despite a markedly elevated 24-h mean plasma cortisol level. These results support the concept that Cushing's disease is associated with a blunting of

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 30 December 2014. at 09:40 For personal use only. No other uses without permission. . All rights reserved.

TABLE 3. Laboratory data before and after normalization of cortisol circadian rhythm Patient L.G. Before Cortisol (jug/dl)


ACTH (pg/ml)


After 4.4



Cortisol secretory episodes/h


Patient C.S. Before


18.3 22





CV (%)





Lowest cortisol value (jug/dl) Metyrapone test Control Day of metyrapone Day after metyrapone





8.4" 11.6 18.7

6.2" 11.8 27.2

" 17-Hydroxycorticosteroids (milligrams per 24 h). CORTISOL HQ per dl 30

JCE & M • 1979 V0U8 • No 5




pretranssphenoidol surgery



20 11 months post - transsphenoidal surgery










FIG. 4. Twenty-four-hour cortisol secretory patterns in patient L.C. before (upper graph) and 11 months after (lower graph) transsphenoidal adenomectomy.

the sensitivity of the CRF-ACTH-secreting axis to cortisol feedback. Recently, Tyrrell and his colleagues reported selective removal of basophilic microadenomas in 17 of 18 patients with Cushing's disease (6). The patient who did not have selective removal required a total hypophysectomy to identify a 1.5-mm basophilic adenoma within the pituitary gland. The resolution of hypercortisolism in 17 of these patients strongly suggests that Cushing's disease is very frequently associated with ACTH-secreting pituitary tumors. One of the questions that remains unanswered is whether these pituitary basophilic adenomas arise de novo or in response to excessive stimulation by the hypothalamic CRF. The results of the present study add further support to

the important findings of Tyrrell et al. (6) by demonstrating that patients with normal polytomes of the sella turcica have small basophilic pituitary tumors that can be identified at the time of transsphenoidal microsurgery. These findings and similar reports by other investigators (12-14) suggest that ACTH-secreting pituitary tumors may be a universal concomitant of Cushing's disease. The reversal of hypercortisolism after transsphenoidal adenomectomy substantiates the role these tumors play in the pathophysiology of this disorder. To gain some insight regarding the underlying etiology of Cushing's disease (hypothalamic vs. pituitary), we studied the circadian pattern of cortisol secretion and metyrapone responsiveness in three of our eight patients after clinical remission had been achieved. Two of the patients showed normalization of their cortisol circadian rhythm and normal metyrapone responsiveness. The third patient showed a failure to normalize her circadian rhythm of cortisol secretion, despite a normal 24-h mean plasma cortisol level. She also had a subnormal metyrapone response. Whether the persistently abnormal cortisol circadian rhythm in this patient implies a continuing hypothalamic abnormality cannot be answered with certainty at this time. It is also possible that it is a manifestation of impaired ACTH reserve or incomplete pituitary tumor removal. If Cushing's syndrome recurs in this patient, it would lend support to either a hypothalamic abnormality or incomplete resection of the tumor. The present study also has defined the 24-h circadian pattern of cortisol secretion in patients with Cushing's disease associated with ACTH-secreting pituitary tumors. The evidence thus far suggests that the vast majority of patients with Cushing's disease have ACTHsecreting pituitary tumors. Whether there are patients with Cushing's disease who do not have pituitary tumors remains to be determined. Removal of these pituitary TABLE 4. Laboratory data for patient L.C. before and after failure of normalization of cortisol circadian rhythm Before Cortisol (jug/dl)


ACTH (pg/ml)


Cortisol secretory episodes/h CV (%)

Lowest cortisol value (jug/dl) Metyrapone test* Control Day of metyrapone Day after metyrapone

0.50 23.9

After" 5.5

Circadian cortisol secretory rhythms in Cushing's disease.

0021 -972X/79/4805-0760$02.00/0 Jouriftl of Clinical Endocrinology and Metabolism Copyright © 1979 by The Endocrine Society Vol. 48, No. 5 Printed in...
626KB Sizes 0 Downloads 0 Views