|Home | About | Journals | Submit | Contact Us | Français|
In patients with multiple endocrine neoplasia type 1 (MEN1), Cushing’s syndrome (CS) from endogenous hypercortisolism can result from pituitary, adrenal, or other endocrine tumors. The purpose of this study was to characterize the range of presentations of CS in a large series of MEN1 patients.
Retrospective review of NIH Clinical Center inpatient records over an approximately 40 year period.
19 patients (8 males, 11 females) with CS and MEN1.
Biochemical, imaging, surgical, and pathological findings.
An etiology was determined for 14 of the 19 patients with CS and MEN1: 11 (79%) had Cushing’s disease (CD) and three (21%) had ACTH-independent CS due to adrenal tumors, frequencies indistinguishable from sporadic CS. Three of 11 MEN1 patients with CD (27%) had additional non-ACTH secreting pituitary microadenomas identified at surgery, an incidence 10-fold higher than in sporadic CD. Ninety-one percent of MEN1 patients with CD were cured after surgery. Two of three MEN1 patients with ACTH-independent CS (67%) had adrenocortical carcinoma. One patient with adrenal cancer and another with adrenal adenoma were cured by unilateral adrenalectomy. No case of ectopic ACTH secretion was identified in our patient cohort. The etiology of CS could not be defined in five patients; in three of these, hypercortisolism appeared to resolve spontaneously.
The tumor multiplicity of MEN1 can be reflected in the anterior pituitary, MEN1-associated ACTH-independent CS may be associated with aggressive adrenocortical disease, and an etiology for CS in MEN1 may be elusive in a substantial minority of patients.
Cushing’s syndrome (CS) from endogenous hypercortisolism can occur sporadically or in the context of familial disease as a result of pituitary or nonpituitary neuroendocrine tumors 1. Among familial endocrine tumor syndromes, multiple endocrine neoplasia type 1 (MEN1) is unique because CS in this setting can result from both ACTH-dependent (Cushing’s disease [CD] and ectopic tumoral ACTH secretion [EAS]) and ACTH-independent causes.
MEN1 is a familial cancer syndrome whose prominent expressions include primary hyperparathyroidism, anterior pituitary tumors, foregut carcinoid tumors, and enteropancreatic endocrine tumors 2. Pituitary tumors are present in about 40% of adult MEN1 patients; 5–10% of such tumors secrete ACTH 3, 4. Functional and particularly non-functional adrenocortical tumors occur frequently, and include benign and malignant tumors causing CS 5. EAS from thymic carcinoids 6, 7 has also been reported in MEN1. MEN1 can include non-endocrine tumors in several tissues 2.
No large series has described the range of presentation and etiologies of CS in MEN1. We report here a retrospective review of 19 patients with CS and MEN1 admitted over four decades to the inpatient wards of the NIH Clinical Center.
By iterative screening of medical records we retrospectively identified patients with CS and MEN1 admitted to the NIH Clinical Center between 1966 and 2009. MEN1 patients were admitted for periodic monitoring or for management of expressions of MEN1. Patients were not specifically recruited or rejected for evaluation of CS. Approximately 120 patients were identified initially by a computer search of discharge diagnoses. Their medical records were individually reviewed to confirm documentation of hypercortisolism and signs and symptoms of CS and MEN1, resulting in the identification of the 19 patients described here. All patients participated in protocols approved by the Investigational Review Boards of the National Institute of Diabetes and Digestive and Kidney Diseases or the Eunice Kennedy Shriver National Institute of Child Health and Human Development, and each gave written consent.
For MEN1 patients initially diagnosed and/or treated for CS outside the NIH Clinical Center, and who then subsequently came to the NIH, all available pertinent outside results were reviewed. For patients evaluated at the NIH, morning electrolytes, glucose, serum cortisol, plasma ACTH, 24-hr urinary excretion of free cortisol (UFC) and/or 17-hydroxycorticosteroid (17OHCS) were measured. The normal range for UFC was 70–300 nmol/d, and the normal range for 17OHCS was 6–28 µmol/d. Additional studies included high-dose dexamethasone suppression testing (HDDST) 8, 9, ovine CRH stimulation test 10, and/or bilateral inferior petrosal sinus sampling (IPSS) before and after administration of CRH 11. A positive result for HDDST was defined as suppression by greater than 90% of UFC or greater than 64% of urinary 17OHCS 8, a positive ovine CRH stimulation test was defined as an ACTH response greater than 35% above baseline at 15 or 30 minutes after CRH injection 10, and a positive central-to-peripheral gradient in bilateral IPSS testing was defined as a ratio of ≥ 3:1 following CRH administration 11. Additional imaging studies for the differential diagnosis of CS, as appropriate for selected patients, included computed tomography (CT) and/or magnetic resonance imaging (MRI), and [131I]-19-iodocholesterol scintigraphy. Some tests were not available or used in the earlier years. Patients received a provisional diagnosis of CD, EAS, or ACTH-independent CS based on the results of dynamic testing. The final diagnosis was based on surgical cure and/or pathological and immunohistochemical examination of tumor obtained at surgery or autopsy.
Continuous data are presented as mean ± S.E.M. Categorical data were analyzed by Fisher’s exact test using two-tailed P values, with statistical analyses performed utilizing Prism software (version 5.0c, GraphPad Software, Inc., La Jolla, CA).
Among approximately 400 MEN1 patients seen at the NIH Clinical Center between 1966 and 2009, nineteen had documented hypercortisolism and/or signs and symptoms of CS (Table 1). Their mean age at diagnosis of CS (36 ± 15.9 yr) was similar to that of unselected CS patients 12. Nevertheless, the presence of three adolescents with CS in our series is striking (Table 1). Fourteen patients (74%) had MEN1 on a familial basis, while 5 (26%) had sporadic disease (Table 1). Germline MEN1 mutational testing results were available on 11 of 19 patients and showed a distribution of truncation and missense mutations across the MEN1 gene similar to those previously reported in MEN1 patients (Table 1, Supplementary Figure 1) 13. Based on biochemical findings, response to surgery and/or pathological and immunohistochemical examination of tumor, the causes of CS were CD (CD group, n = 11) or ACTH-independent CS due to adrenal tumors (CA group, n = 3). In five patients, no clear etiology (CX group) could be determined (Table 1, and below). No patient was proven to have EAS. Among the 14 patients with CS for whom an etiology could be identified, the frequencies of CD (11 cases; 79%) and ACTH-independent CS (3 cases; 21%) resembled those reported in non-familial CS 12, 14.
Eighteen of 19 patients had elevated UFC or 17-OHCS at the time of initial presentation (Table 2).
Sixteen of the 19 patients had one or more dynamic test results available (Table 2). Among the CD group with testing results, all had positive CRH-stimulation tests 10 (n=8), suppressed their serum cortisol and/or UFC or urinary 17OHCS in response to high-dose dexamethasone administration 8, 9 (n=8), and showed central-to-peripheral gradients of ACTH during IPSS testing (n=9) (Table 2). Within the CA group, one patient failed to respond to either CRH or HDDST, and another did not respond to HDDST but responded to CRH test and had a central-to-peripheral ACTH gradient during IPSS testing (Table 2). Interestingly this latter patient (no. 13), who did not appear to have cyclic CS, underwent unsuccessful transsphenoidal pituitary surgery (TSS) before she was cured by recognition and removal of an adrenal adenoma (see below). All five of the patients in the CX group underwent at least one dynamic test (Table 2).
Sixteen of the 19 patients with MEN1 and CS had pituitary and/or adrenal imaging results (Table 3). Among 10 patients in the CD group for whom preoperative pituitary imaging was available, six had pituitary micro- or macroadenomas or pituitary enlargement documented, while four had normal studies. Two CA patients had normal pituitary imaging and large unilateral adrenal masses. Four of the five patients in the CX group had one or more relevant imaging result available (Table 3).
Ten of 11 patients in the CD group had TSS (Table 4). One or more pituitary adenoma was identified at surgery in eight patients. In two cases no tumor was found, but these patients had biochemical and clinical remission after TSS and hemi-hypophysectomy (patients 2 and 8). Three patients of the 11 with CD (27%) had two separate pituitary microadenomas identified at surgery (patients 1, 3, and 10). This incidence of multiple pituitary adenomas is significantly higher than both that reported previously in a large series of CD cases (11 out of 658 total, excluding two patients with MEN1; ~2%) 15 (P < 0.001) and that found in a series of unselected sporadic pituitary adenomas (3 out of 116; 2.6%) 16 (P < 0.01). Of note, four of the CD patients had macroadenomas or invasive tumors at surgery and another developed Nelson’s syndrome after adrenalectomy.
Seven of the nine patients in the CD group for whom surgical pathology reports were available had at least one immunoreactive microadenoma identified (Table 4). An ACTH-positive tumor was documented in four. Among the three patients with two adenomas, one had two distinct prolactin (PRL)-staining tumors (patient 3), while two had an ACTH-positive tumor and a second distinct PRL or PRL/growth hormone immunoreactive tumor (patients 1 and 10 respectively). Nine of the 10 patients with CD were cured after TSS (Table 4).
The three CA patients had adrenal cortical tumors identified by surgery or autopsy. Two (67%) had adrenocortical carcinoma (ACC) based on pathologic identification of vascular invasion (patient 12) or pulmonary metastases (patient 14). The incidences of ACC among the 14 MEN1 patients with an identifiable etiology and among the subset with ACTH-independent CS were similar to those reported in large series of unselected CS patients 12, 14.
Three patients among five in the CX group underwent surgery directed at CS (Table 4). Two patients had failed TSS (patients 16 and 19) without tumor visualization at surgery. Although patient 19 had a hemi-hypophysectomy with a 1 mm ACTH-immunoreactive microadenoma subsequently reported in the excised tissue, hypercortisolism persisted during 24 months of follow-up. By contrast, patient 16 had spontaneous resolution of hypercortisolism between four and 14 months after surgery (Supplemental Table 1). Patient 15 had bilateral ADX at which time an adrenocortical tumor and contralateral pheochromocytoma were removed. Although this patient was cured after ADX it is unclear whether the hypercortisolism was ACTH-independent and resulting from the adrenocortical tumor, or due to ectopic release of ACTH from the pheochromocytoma. ACTH staining of the pheochromocytoma was not performed at the time of surgery (1971) and the archival specimen was not retrievable.
The remaining two patients in the CX group had spontaneous resolution of hypercortisolism without any surgical intervention.
CS is an uncommon manifestation of MEN-1 that can be caused by both ACTH-dependent and independent disorders. Although none of the 19 patients reported here had clear-cut ectopic ACTH secretion, the proportion of CD and adrenal etiologies was similar to that reported in non-familial cases, underscoring the need to consider all etiologies of CS in patients with MEN1.
The high incidence of multiple pituitary adenomas discovered at surgery in MEN1 patients with CD reported here must be seen in the context of other “multiplicities” that characterize this familial tumor syndrome. Not only are multiple endocrine organs affected by the tendency to neoplasia, but also within susceptible tissues the tumors are frequently multiple. Thus it perhaps is not surprising that pituitary tumor multifocality would be observed, since the anterior pituitary is frequently affected in MEN1 3.
Why then, was the frequent multifocality of pituitary tumors in MEN1 not recognized in a large series of pituitary adenomas in MEN1 patients by Vergès et al 3? These authors studied a cohort of 324 MEN1 patients including 136 with pituitary disease and reported no multiple pituitary tumors 3. In contrast to the microadenomas found in the majority of MEN1 patients with CD described here, 85% of pituitary tumors in the series reported by Vergès et al were macroadenomas 3, 17. The presence of large macroadenomas in the Vergès series may have obscured detection of coexisting microadenomas during neurosurgery 3.
The frequency of ACC among the MEN1 patients described here, though striking, did not reach statistical significance compared to surveys of largely sporadic CS 12, 14, due in part to the small numbers in our series. ACC has been previously reported in MEN1 patients with CS 5, 18–20.
Our inability to determine the etiology of CS in five of the 19 MEN1 patients in our series is puzzling, since such a large category of diagnostic unknowns was not similarly recognized in two large contemporary series of sporadic CS 12, 14. One possibility is that not all patients underwent IPSS, which is the best test for the differential diagnosis (11). Patient 15 underwent bilateral adrenalectomy in 1971 prior to the development of IPSS testing, whereas the hypercortisolism of patients 16 and 17 resolved prior to scheduled IPSS. Since patient 16 had resolution of hypercortisolism between four and 14 months after TSS (Supplemental Table 1), and delayed remission of Cushing’s disease following TSS has been reported in a subset of cases 21, 22, it is possible that patient 16 had CD. Even though EAS as a cause of CS has been reported rarely in MEN1 6, 7, our failure to identify EAS as the cause of CS in any of our 19 patients is consistent with the absence of any MEN1 patient among the 90 cases of EAS reported from our institution in a 20-year retrospective review 23. Spontaneous resolution of CS may have contributed to our inability to identify the cause of CS in at least two of the five patients in our CX group (Table 4). Since periodic or cyclic CS occurs with all etiologies of CS 24, the spontaneous resolution of CS does not in itself offer any diagnostic clue.
We thank Ms. Michelle Hendery of the NIH Clinical Center for her assistance with patient database searching. We acknowledge the expert care of patients in this study by our present and former colleagues, including Drs. Allen M. Spiegel, Lee S. Weinstein, Michael T. Collins, Monica C. Skarulis, Carmen M. Mateo, Robert T. Jensen, Stephen A. Wank, Constantine A. Stratakis, George P. Chrousos, Gordon B. Cutler, Jr., D. Lynn Loriaux, Electron Kebebew, Steven K. Libutti, H. Richard Alexander, Edward H. Oldfield, and the late John L. Doppman (1928–2000). We are furthermore grateful to our patients, and to Mr. Craig Cochran and the nurses and fellows of the endocrinology wards at the NIH Clinical Center for their excellent patient care. The intramural research programs of Eunice Kennedy Shriver National Institute of Child Health and Human Development and the National Institute of Diabetes and Digestive and Kidney Diseases funded this study.
None of the authors has a conflict of interest to declare.