|
|
| |||||||||||||||||||||||||||||||
Background and Methods Cushing's syndrome is rare in children and adolescents. We analyzed the clinical presentation, diagnostic evaluation, and treatment of 59 patients with Cushing's syndrome between the ages of 4 and 20 years who were admitted to the National Institutes of Health during the period from 1982 to 1992. The cause of hypercortisolism was identified by low- and high-dose dexamethasone suppression tests, the ovine corticotropin-releasing hormone (CRH) stimulation test, imaging studies, and bilateral sampling of the inferior petrosal sinuses combined with administration of CRH.
Results Fifty patients had Cushing's disease, six had primary adrenal disease, and three had ectopic corticotropin secretion. The initial signs were excessive weight gain in 90 percent of the patients and growth retardation in 83 percent. Most patients (81 percent) had normal bone age at the time of diagnosis. Forty-seven percent had hypertension, whereas only 19 percent had mental or behavioral problems. The high-dose dexamethasone suppression test and the CRH stimulation test identified 68 and 80 percent, respectively, of the patients with Cushing's disease. Magnetic resonance imaging of the pituitary indicated the presence of tumor in 52 percent of the patients with pituitary adenomas. The maximal central-to-peripheral ratio of plasma corticotropin during sampling of the interior petrosal sinuses was
Conclusions Weight gain and growth retardation are common clinical characteristics of Cushing's syndrome in children and adolescents. Diagnostic evaluation of such patients with CRH stimulation alone and combined with inferior petrosal sinus sampling and imaging studies is accurate, and therapy is usually successful.
2.5 in all the patients with Cushing's disease and <2.5 in those with ectopic corticotropin secretion. Remission of hypercortisolism was achieved in 48 of the 49 patients who underwent transsphenoidal surgery for Cushing's disease, in all 6 of the patients who underwent adrenalectomy for primary adrenal disease, and in the 2 patients in whom the ectopic source of corticotropin could be identified.
Methods
Subjects
Since 1982, 59 children and adolescents (37 girls and 22 boys; mean age, 14 years; range, 4 to 20) admitted to the National Institutes of Health (NIH) Clinical Center received a diagnosis of Cushing's syndrome. The records of these patients, systematically completed by endocrine fellows as part of a protocol, were reviewed retrospectively. The diagnosis was based on the history, physical examination, and laboratory evidence of cortisol hypersecretion. The cause of hypercortisolism was identified by the standard suppression test with low- and high-dose dexamethasone, the stimulation test with corticotropin-releasing hormone (CRH), imaging studies (computed tomography [CT] or magnetic resonance imaging [MRI] of the pituitary, adrenal glands, or chest), and sampling of the inferior petrosal sinuses before and after administration of CRH. The final diagnosis was established by surgical and pathological findings and cure after the appropriate treatment. The clinical characteristics of the patients, according to the causes of Cushing's syndome, are summarized in Table 1. All patients or their parents gave informed consent for the studies.
|
Diagnostic Evaluation
Base-Line Studies
For each patient, 5 to 15 measurements of base-line 24-hour urinary cortisol, 17-hydroxycorticosteroid, and creatinine excretion were averaged to determine the mean urinary cortisol excretion per square meter of body-surface area and the mean 17-hydroxycorticosteroid excretion per gram of creatinine13,14. Blood samples for the determination of plasma cortisol concentrations were obtained at 6 a.m., 6:30 a.m., 7 a.m., 7:30 a.m., and 8 a.m. and at 11 p.m., 11:30 p.m., midnight, 12:30 a.m., and 1 a.m. The averages of the morning and evening values were used to evaluate the circadian rhythmicity of cortisol secretion. At least two blood samples were drawn in the morning (7 to 9 a.m.) for determination of plasma corticotropin concentrations.
Dexamethasone Suppression Test
The standard suppression test with low- and high-dose dexamethasone was performed as previously described,15 with dexamethasone administered according to body weight (30 and 120 µg per kilogram per day, respectively)2. The mean values for urinary cortisol and 17-hydroxycorticosteroid at base line (two days) were compared with the values on the second day of high-dose dexamethasone administration. A response was considered positive if the urinary 17-hydroxycorticosteroid or cortisol excretion was reduced by at least 64 or 90 percent, respectively, below the mean base-line value10.
CRH Stimulation Test
Ovine CRH (Bachem, Torrance, Calif.) was prepared and administered as previously described16,17. Blood samples for determination of plasma corticotropin and cortisol concentrations were obtained 15 minutes and 1 minute before and 5, 15, 30, 60, 90, and 120 minutes after the administration of CRH. The mean values of plasma corticotropin and cortisol concentrations before and after CRH administration were compared. A response was considered positive if the plasma corticotropin or cortisol value was increased by at least 34 or 20 percent, respectively, over the mean base-line value11.
Imaging Studies
MRI scans of the pituitary gland, with or without gadolinium, CT scans of the adrenal glands, and chest x-ray films were obtained in all patients. CT and MRI scans of the chest and abdomen were obtained if the suspected cause of the disorder was a nonpituitary tumor secreting corticotropin.
Bilateral Sampling of the Inferior Petrosal Sinuses
Bilateral sampling of the inferior petrosal sinuses was combined with the administration of CRH, as previously described18. Blood samples were obtained 3 minutes and 1 minute before and 3, 5, and 10 minutes after the administration of CRH. The plasma corticotropin concentrations before and after the administration of CRH were used to calculate the ratio of the corticotropin concentration in each of the inferior petrosal sinuses to the concurrent concentration in peripheral-blood plasma (central-to-peripheral ratio). A central-to-peripheral ratio 2.0 before or 3.0 after the administration of CRH was used to identify the pituitary gland as the source of corticotropin, and a ratio 1.4 between the two sinuses (lateralization ratio) was used to predict the site of the tumor in the patients with Cushing's disease12.
Outcome of Surgery
Patients with corticotropin-secreting pituitary adenomas, ectopic tumors, and primary adrenal disease were treated with transsphenoidal surgery, surgical tumor excision, and bilateral or unilateral adrenalectomy, respectively. Patients were considered cured if urinary cortisol values were <10 µg per 24 hours (<28 nmol per day) and morning (7 to 9 a.m.) plasma cortisol values were <1 µg per deciliter (<28 nmol per liter) the third day after surgery (36 hours after temporary discontinuation of dexamethasone). Postoperative recovery of the hypothalamic-pituitary-adrenal axis was evaluated periodically by serial tests with 0.25 mg of intravenous corticotropin 1-24 (Cortrosyn; Organon, West Orange, N.J.) and was confirmed by a plasma cortisol value 18 µg per deciliter (500 nmol per liter) 60 minutes after the administration of corticotropin.
Patients who underwent transsphenoidal surgery were evaluated for hypopituitarism. Central hypothyroidism was diagnosed if plasma free thyroxine values were 
1.0 ng per deciliter (13 pmol per liter). Growth hormone deficiency was diagnosed if plasma growth hormone values were <5 µg per liter after the administration of both arginine and levodopa and if the growth rate was <4 cm per year. Hypogonadism was suspected clinically if pubertal development was delayed, and the diagnosis was confirmed by low plasma sex hormone concentrations and poor plasma gonadotropin responses to gonadotropin-releasing hormone. Diabetes insipidus was suspected if the volume of urine exceeded 30 ml per kilogram of body weight per day and the plasma sodium concentration was above 145 mmol per liter. Measurements of 24-hour urinary cortisol excretion were used to evaluate the patients postoperatively for the recurrence of hypercortisolism.
Hormone Assays
Urinary cortisol excretion was measured by direct radioimmunoassay19. The intraassay and interassay coefficients of variation were 5 and 10 percent, respectively. Normal values do not vary with age if corrected for body-surface area14. Urinary 17-hydroxycorticosteroid excretion was measured by a modification of the colorimetric method of Porter and Silber20. The intraassay and interassay coefficients of variation were 6 and 11 percent, respectively. Normal values do not vary with age if corrected per gram of creatinine excretion. Plasma corticotropin21 and cortisol22 were measured by radioimmunoassay, as previously described. The detection limits for the corticotropin and cortisol assays were 3 to 10 pg per milliliter (0.7 to 2.2 pmol per liter) and 0.2 to 1.0 µg per deciliter (5.5 to 28 nmol per liter), respectively. The intraassay and interassay coefficients of variation were 8 and 15 percent, respectively, for corticotropin and 4.6 and 6.0 percent, respectively, for cortisol.
Results
Clinical Profile and Diagnosis
In most of the patients the onset of the disease was insidious, taking months or years for all the symptoms to develop (mean [±SD], 3 ±2 years; range, 3 months to 7 years) (Table 1). The patients' clinical characteristics are shown in Table 2, and the data on their growth in Figure 1 and Table 1. The height of the prepubertal patients was more impaired than that of the pubertal patients. School performance was normal in all 20 patients for whom information was available. All the patients had biochemically confirmed hypercortisolism and lacked diurnal variation in plasma cortisol concentrations (Table 3). Periodicity (periods of hypercortisolism followed by periods of eucortisolism) was observed in only two patients: one with micronodular adrenal disease and one with ectopic corticotropin secretion23.
|
|
|
Base-Line Corticotropin Concentrations
The morning plasma corticotropin concentrations ranged from 10 to 195 pg per milliliter (2 to 43 pmol per liter) (mean, 58 pg per milliliter [13 pmol per liter]) in the patients with Cushing's disease, from 3 to 14 pg per milliliter (0.7 to 3.0 pmol per liter) (mean, 6 pg per milliliter [1.3 pmol per liter]) in the patients with primary adrenal disease, and from 156 to 1340 pg per milliliter (34 to 295 pmol per liter) (mean, 628 pg per milliliter [138 pmol per liter]) in the patients with ectopic corticotropin secretion (Table 3).
Dexamethasone Suppression Test
Forty-six patients underwent the dexamethasone suppression test (Figure 2). On the second day of administration of high-dose dexamethasone, 25 of 37 patients with pituitary adenomas (68 percent) had a decrease in urinary 17-hydroxycorticosteroid excretion 64 percent below the mean base-line value; 19 of 28 patients with pituitary adenomas (68 percent) had a decrease in urinary cortisol excretion 90 percent below the mean base-line values. None of the five patients with ectopic corticotropin secretion or primary adrenal disease had a decrease in urinary 17-hydroxycorticosteroid or cortisol excretion. All four patients with bilateral micronodular adrenal hyperplasia had an increase in urinary 17-hydroxycorticosteroid and cortisol excretion during the second day of the high-dose dexamethasone test.
|
Fifty-one patients underwent the CRH test (Figure 3). Thirty of 40 patients (75 percent) with pituitary adenomas had an increase in plasma corticotropin 34 percent after the administration of CRH, when the mean base-line values were compared with the means of all values after CRH administration. None of the seven patients with nonpituitary Cushing's syndrome had a response to CRH.
|
20 percent after the administration of CRH, when the mean base-line values were compared with the means of all values after CRH administration. Of the six patients with nonpituitary Cushing's syndrome, one, who had an adrenal carcinoma, had a 22 percent increase in plasma cortisol after the administration of CRH. Imaging Studies
MRI scans of the pituitary with gadolinium revealed pituitary tumors in 26 of the 50 patients with Cushing's disease (52 percent). The results were equivocal in 11 patients (22 percent) and normal in 13 (26 percent). CT imaging of the abdomen revealed normal-size adrenal glands in 22 of the 50 patients (44 percent) and enlarged adrenal glands in 28 (56 percent). CT scans of the adrenal glands were diagnostic in both patients with adrenal carcinoma. Among the four patients with bilateral micronodular adrenal hyperplasia, the CT scans showed bilateral or unilateral nodularity with normal-size glands in two patients, and normal glands in the other two. The CT images in the three patients with ectopic corticotropin secretion revealed normal adrenal glands in one and hyperplastic glands in two. Chest x-ray films and CT scans revealed tumors in two of these patients.
Bilateral Sampling of the Inferior Petrosal Sinuses
Bilateral sampling of the inferior petrosal sinuses was performed successfully in all 50 patients in whom it was attempted, with no side effects. The two youngest patients were five years old. In four patients the results were uninterpretable or inconsistent with those of other tests, and all four underwent repeated sampling, with diagnostic results.
Forty-one of 43 patients with Cushing's disease (95 percent) had a maximal central-to-peripheral ratio 2 for base-line plasma corticotropin concentrations (Figure 4),12 as compared with none of the 7 patients with ectopic corticotropin secretion or primary adrenal disease in whom the test was performed. After the administration of CRH, 42 of the 43 patients with Cushing's disease (98 percent) had a maximal central-to-peripheral ratio 3,12 as compared with none of the 7 other patients. All the patients with Cushing's disease had a maximal central-to-peripheral ratio 2.5 before and after the administration of CRH, whereas all the other patients had a maximal central-to-peripheral ratio 2.2 before and after CRH administration.
|
1.4 between the plasma corticotropin concentrations in the two sinuses correctly predicted the lateralization of the microadenoma in 67 and 76 percent of the patients before and after CRH administration, respectively12. Treatment
Forty-nine of the 50 patients with Cushing's disease underwent transsphenoidal surgery. The operation was a selective adenomectomy in 39 patients, 10 of whom had previously undergone pituitary surgery, and a total hypophysectomy in 1, a patient previously operated on twice. Nine patients (including one with prior pituitary surgery) in whom tumors were not identified at surgery underwent a hemihypophysectomy. One patient, referred after three previous operations, was treated with radiation (50.4 Gy in 6 weeks) and mitotane, which was discontinued 2 1/2 years later, when urinary cortisol excretion was normal.
All four patients with bilateral micronodular adrenal hyperplasia underwent bilateral adrenalectomy. The two patients with adrenal carcinomas were treated with surgical excision and then with mitotane.
In two of the three patients with ectopic corticotropin secretion, the tumors were identified and surgically removed. The third patient underwent thymectomy on the basis of a corticotropin gradient on thymic-vein sampling. No tumor was found, and a bilateral adrenalectomy was performed one month later. This patient presumably had an intrathoracic tumor that eluded detection.
Pathological Studies
The size of the 39 pituitary adenomas ranged from 1.5 to 11 mm; only 2 were macroadenomas (10 mm and 11 mm). Thirty-three of the adenomas (85 percent) were composed of cells that stained with anti-corticotropin antibodies. A 3-mm pituitary adenoma was identified in the excised hemipituitary in one of the nine patients (11 percent) who underwent a hemihypophysectomy. Among the six patients with adrenal disease, four had bilateral pigmented micronodular hyperplasia and two had anaplastic adrenal carcinomas. Among the three patients with ectopic corticotropin secretion, one had a bronchial carcinoid, and one a thymic carcinoid; both were positive for corticotropin on immunostaining.
Outcome
On the third postoperative day, 57 patients had urinary cortisol values <10 µg per 24 hours (<28 nmol per day) and a morning plasma cortisol value <1 µg per deciliter (<28 nmol per liter); these patients were considered cured. They included 48 of the 49 patients (98 percent) with Cushing's disease who underwent transsphenoidal surgery and all 37 patients who had not had previous pituitary surgery. In one patient surgery failed but treatment with mitotane was successful. All patients with pituitary Cushing's disease after surgery needed hydrocortisone therapy for a minimum of six months.
The follow-up period ranged from 5 to 60 months (mean, 22). Fifty-six patients had no recurrence of disease. In 2 of the 37 patients with Cushing's disease who had not had previous surgery (5 percent), both of whom had tumors that had invaded the wall of the cavernous sinus, hypercortisolism recurred 14 and 24 months after surgery. One of the two patients underwent successful repeated transsphenoidal adenomectomy, and the other underwent radiation therapy. One of the 12 patients who had repeated transsphenoidal surgery at the NIH (8 percent) had recurrent hypercortisolism four months after surgery and received radiation therapy. Thus, all three patients with recurrent or persistent hypercortisolism ultimately had a remission.
One year after surgery, most of the patients had decreased body weight and body mass, and their height and growth velocity had increased (Table 1).
Among the 37 patients with Cushing's disease first operated on at the NIH, 7 (19 percent) had one or more endocrine deficiencies: central hypothyroidism (in 6 patients), diabetes insipidus (in 3), growth hormone deficiency (in 3), and central hypogonadism (in 1). The diabetes insipidus resolved in all three patients 6 to 20 months after surgery.
All four patients with bilateral micronodular adrenal hyperplasia had a remission of hypercortisolism. The two patients with adrenal carcinomas had hepatic and pulmonary metastases 9 to 12 months after surgery and died 12 and 30 months, respectively, after the onset of Cushing's syndrome.
Discussion
Weight gain, obesity, and facial plethora were the prevailing signs of Cushing's syndrome in our young patients, and the frequency of these symptoms was similar to that reported in adults1,2,4,5,24,25. Growth retardation was a common manifestation in patients who were not fully grown. About 40 percent of the children whose illness had started during the first decade of life presented with premature sexual development, presumably caused by increased secretion of adrenal androgens. On the other hand, 3 percent of the children who were older than 13 years (girls) or 14 years (boys) at the time of the onset of Cushing's syndrome were prepubertal, probably because of hypogonadism induced by hypercortisolism24,25. Contrary to previous reports of delayed bone age in patients with hypercortisolism,3,25 bone age was consistent with chronologic age in 81 percent of our patients. Bone age was accelerated in 8 percent of the patients and delayed in 11 percent, and was clinically correlated with early and delayed sexual development, respectively. Apparently, the bone age of the patients reflected the combined effects of cortisol, which should have had an inhibitory effect, and adrenal androgens and gonadal steroids, which should have had a stimulatory effect. Mental changes (emotional lability, irritability, or depression), muscle weakness, and sleep disturbances were reported less often in our patients than in adults, and unlike the job performance of adults with Cushing's syndrome, which is reported to be poor, the school performance of our children was satisfactory1,4,5,26.
Elevated values for 24-hour urinary cortisol or 17-hydroxycorticosteroid excretion were equally reliable for the diagnosis of hypercortisolism. The high-dose dexamethasone suppression test and the CRH stimulation test identified 68 and 80 percent, respectively, of patients with Cushing's disease. Since the CRH test can be completed in two hours, it is more convenient and less costly than the dexamethasone suppression test and at least as reliable. The diagnostic accuracy of pituitary MRI in our patients (52 percent) was equivalent to that of MRI in adults27. Bilateral sampling of the inferior petrosal sinuses was the best test for the differential diagnosis of corticotropin-dependent Cushing's syndrome, with a diagnostic accuracy of 100 percent.
Transsphenoidal adenomectomy is the treatment of choice for children with Cushing's disease, as it is in adults28,29,30. Generally, when tumors less than 10 mm in diameter (microadenomas) are identified, the results are excellent, with remission of the disease in 85 percent of patients. In our series 98 percent of the patients who underwent transsphenoidal surgery, including all those who had not had prior pituitary surgery, entered remission; 5 percent had a recurrence. We prefer repeated transsphenoidal surgery to radiotherapy in patients who have relapses after surgery for Cushing's disease31. Radiation, with or without concomitant mitotane treatment, is an alternative, however, especially if dural or cavernous sinus invasion was observed during the first procedure32. Bilateral adrenalectomy may be necessary in patients with Cushing's disease in whom both pituitary surgery and pituitary radiation with mitotane treatment have failed.
Referral of patients to the NIH introduced a selection bias. For example, patients with Cushing's disease with normal MRI or CT scans of the pituitary were referred for more detailed diagnostic assessment, including CRH testing and bilateral sampling of the inferior petrosal sinuses, and patients in whom previous transsphenoidal surgery had been unsuccessful were referred for reevaluation and reoperation. In addition, mildly affected patients with disease that was difficult to diagnose or those with equivocal test results were referred for confirmation of the diagnosis and differential diagnosis. Despite the selection bias, the data from this large series provide useful information on the presentation, diagnosis, and treatment of Cushing's syndrome in children and adolescents.
On the basis of our results, we recommend the following diagnostic scheme in children and adolescents with Cushing's syndrome. First, the presence of hypercortisolism should be established by demonstrating increased urinary cortisol or 17-hydroxycorticosteroid excretion on several occasions. Second, plasma corticotropin should be measured and a CRH test performed to distinguish between corticotropin-dependent and corticotropin-independent Cushing's syndrome, and pituitary MRI scans should be obtained in all patients with the former type of disorder. In the case of a positive CRH test and an unequivocally positive pituitary MRI scan, transsphenoidal surgery is indicated. If the pituitary MRI scan is negative or equivocal, bilateral sampling of the inferior petrosal sinuses should be performed. Patients with CRH test results that suggest ectopic corticotropin secretion and negative pituitary MRI scans should also undergo bilateral sampling of the inferior petrosal sinuses and chest and abdominal CT or MRI scans. If an ectopic corticotropin-secreting tumor is identified, it should be excised. If basal plasma corticotropin values suggest corticotropin-independent Cushing's syndrome, CT or MRI scanning of the adrenal glands should be performed. Adrenal adenomas and carcinomas, which are usually easy to detect on scans, require surgical excision; bilateral micronodular adrenal hyperplasia requires bilateral adrenalectomy.
Supported in part by a grant (to Dr. Magiakou) from the Alexander S. Onassis Public Benefit Foundation.
We are indebted to Drs. William Travis and David Katz for the pathological evaluations and to the medical-staff fellows and nursing staffs of the Developmental Endocrinology Branch, National Institute of Child Health and Human Development, the Surgical Neurology Branch, National Institute of Neurological Disorders and Stroke, and the Diagnostic Radiology Department of the Clinical Center, National Institutes of Health, who provided superb care to the patients.
Source Information
From the Developmental Endocrinology Branch, National Institute of Child Health and Human Development (M.A.M., G.M., M.T.G., G.B.C., L.K.N., G.P.C.); the Surgical Neurology Branch, National Institute of Neurological Disorders and Stroke (E.H.O.); and the Diagnostic Radiology Department, National Institutes of Health (J.L.D.) -- all in Bethesda, Md.
Address reprint requests to Dr. Magiakou at NICHD/DEB Bldg. 10, Rm. 10N262, 9000 Rockville Pike, Bethesda, MD 20892.
References
| |||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
