| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
From The Clinical Research Centers |
Department of Pediatrics, New York-Presbyterian Hospital, New York Weill Cornell Center (R.S.N., G.K., T.L., M.I.N.), New York, New York 10021; and Department of Human Biological Chemistry and Genetics (D.C., J.A., E.B.T.), University of Texas Medical Branch (UTMB), Galveston, Texas
Address correspondence and requests for reprints to: Maria I. New, M.D., Professor and Chairman, Department of Pediatrics, Chief, Division of Pediatric Endocrinology, Harold and Percy Uris Professor of Pediatric Endocrinology and Metabolism, New York-Presbyterian Hospital, 525 East 68th Street, Room M-622, New York, New York 10021. E-mail: minew{at}mail.med.cornell.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsCase HistoryResultsDiscussionConclusionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsCase HistoryResultsDiscussionConclusionReferences |
|---|
In this study, we describe a girl who developed Cushingoid features in peripuberty and was eucortisolemic. There was no history of exogenous administration of topical, inhaled, or systemic glucocorticoids. She was previously shown to have markedly elevated lymphocyte glucocorticoid receptors (GR) with normal binding affinity to dexamethasone (1), which was speculated to be a potential cause of her phenotype. However, serum cortisol levels were not decreased, and her hypothalamic-pituitary-adrenal (HPA) axis was intermittently dysregulated. This is unlike the intact HPA axis observed in the condition of differential response to thyroid hormones at peripheral vs. hypothalamic-pituitary tissue. Therefore, normal GR numbers at the pituitary, vs. elevated numbers peripherally, may not explain her condition entirely. Nonetheless, she had variable tissue responsivity to glucocorticoids and presented at age 10.8 yr with centripetal obesity, moon facies, buffalo hump, and purple striae, but without easy bruising or statural stunting, which are two cardinal signs of Cushing’s disease in childhood. At 11.5 yr she suffered a compression fracture of the L1 vertebra. The antiglucocorticoid drug mifepristone (RU486) was subsequently administered in doses of 400 mg/day. Striae gradually resolved, and her buffalo hump, weight gain, and distribution improved in response to RU486, despite the up-regulation of the HPA axis and raised cortisol levels. Bone mineral density (BMD) did not improve. Once RU486 was introduced, however, a consistent correlation over time between the Cushingoid features and GR sites per cell was no longer observed. There was a correlation, however, between the number of GR sites and the administration of RU486. Ultimately, her Cushingoid phenotype proved to be transient. Although the precise etiology remains unknown, it is likely to involve the postreceptor transcription apparatus critical to glucocorticoid action.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsCase HistoryResultsDiscussionConclusionReferences |
|---|
ACTH plasma levels were determined commercially by the Nichols Institute, initially as immunoradiometric assay, and most recently using a highly sensitive immuno-chemiluminescence assay. Corticotropin-releasing factor (CRF) test was performed by administering ovine CRF iv over 1 min (5); baseline and subsequent ACTH and/or cortisol levels were determined at 15, 30, 60, 90, 120, and 180 min after CRF injection.
The GR assays were performed on freshly collected, anticoagulated peripheral blood lymphocytes. Blood was put on ice and shipped overnight to UTMB (Galveston, TX). Within 24 h, the peripheral lymphocytes were separated on Ficoll-Hypaque gradients and assayed for GR binding sites by competitive binding assay, as described previously (6), with minor modifications. This whole cell-binding assay has been performed in our laboratory for 2 decades, producing consistent results with the CEM cell line. During this study, periodic assays on CEM cells validated the performance of the assay. Data for specific compatible binding was obtained at up to nine different concentrations of [3H]dexamethasone (2.5 nM to 175 nM) in each assay. GR affinity, expressed as apparent dissociation constant (Kd), and the number of specific dexamethasone binding sites per cell were derived via Scatchard plot analysis.
GR gene and promotor usage analysis
Using nine primers spanning the entire GR coding region as described (7), GR complementary DNA of the human GR was obtained after an initial reverse transcriptase reaction from total RNA derived from the patient’s lymphocytes and amplification by the PCR method. The resultant GR gene was sequenced in the UTMB analytical laboratory and found to be identical to that of the published normal GR gene (8). Quantitative PCR (using increasing amounts of competitor transcript of a deliberately mutated human GR) (9), was used to evaluate the level of GR messenger RNA (mRNA) isolated from the same lymphocytes used for the binding assays from our patient, and they were compared with a normal, age-matched female. The method was validated by observing that GR mRNA levels increase 3.4-fold after exposing the CEM-C7 human leukemic cell line to 1 mM dexamethasone. GR mRNA levels are known to be induced by glucocorticoids in this cell line (10, 11). Promoter usage was checked via PCR [sequence kindly provided by Dr. Jeff Harmon (Department of Pharmacology, Uniformed Services University of the Health Sciences, Bethesda, MD)], using a common 3' primer (between nucleotides 394 and 414 in exon 2) and four different 5' primers from noncoding exons (1A, 1B, 1C, and 1D). Both strands were sequenced. All sequencing was carried out in the UTMB DNA sequencing laboratory using an PE Applied Biosystems Sequencer (PE Applied Biosystems, Foster City, CA).
BMD (g/cm2) was assessed via dual-energy x-ray absorptiometry (Lunar Corp., Madison, WI) at the spine and hip. Normal age-appropriate values were taken from published data on a large Australian cohort, which is of a similar ethnic background as our patient (12).
|
Case History |
|---|
|
TopAbstractIntroductionMaterials and MethodsCase HistoryResultsDiscussionConclusionReferences |
|---|
) or stunting of
statural growth (Fig. 2). She had a mild
learning disability. Physical examination revealed a normal blood
pressure. The patient was in early puberty with breasts at Tanner stage
III (B-III) and pubic hair at stage II (PH-II). Initial hormonal
findings demonstrated that the patient was eucortisolemic, with an
intact circadian rhythm of cortisol, normal 24-h urinary free cortisol,
and an intact pituitary-adrenal axis as assessed by metyrapone test and
low-dose dexamethasone suppression. Head magnetic resonance imaging and
abdominal imaging were normal. An oral glucose tolerance test revealed
marked insulin resistance (peak insulin >400 µU/mL). Bone age was 11
yr, and there was mild osteoporosis noted. As a child, the patient had
bilateral wrist fractures at separate times and a finger fracture, all
secondary to trauma.
|
|
), revealing mild vertebral osteopenia.
BMD at the femurs showed mild to moderate osteopenia. There was no
statural stunting or detectable abnormality in bone formation,
breakdown, or bone metabolism. Tests for metabolic or genetic bone
diseases were normal, including serum PTH, alkaline phosphatase,
calcium, phosphorus, and responses to calcium challenge, urinary
excretion of calcium, phosphorus, and hydroxyproline. Osteogenesis
imperfecta was ruled out specifically by studies on skin fibroblasts.
Height velocity (6.34 cm/yr) and pubertal advancement were normal
(B-III/IV, PH-III), whereas the patient gained an additional 6.2 kg
(8.7 kg/yr), as shown in the growth curve (Fig. 2
). Urinary free
cortisol levels were not elevated (<70 µg/m2·24
h).
|
). A 3-week hiatus in therapy caused
noticeable worsening in the striae and a 1.4-kg weight gain. Resuming
RU486 caused total disappearance of violaceous striae and a mild weight
reduction. At age 14 yr, 3 months (Table 1), therapy was withheld owing
to persistent vaginal bleeding. Within 2 weeks the striae worsened as
she gained 2.1 kg, and after 5–6 weeks the striae returned to their
initial state. Oral contraceptive therapy (OCP) was instituted after an
episode of heavy vaginal bleeding. Two menstrual cycles later, OCP
therapy was discontinued by the parents following improvement; the
vaginal bleeding then resumed. A new vertebral compression fracture
occurred that prompted another course of RU486. At age 15 yr, 6 months,
after 6 months of RU486 (400 mg/day), striae improved markedly, weight
remained stable, and the buffalo hump became softer and smaller in
size. Again, the patient experienced persistent vaginal bleeding and
had one episode of dysfunctional uterine bleeding. Because the patient
could not tolerate pelvic sonography, magnetic resonance imaging was
carried out, showing a distended uterus filled with amorphous material
that extended through the cervix into the proximal vagina (R. S.
Newfield et al., submitted for publication). As anticipated,
while still receiving RU486 (400 mg/day), no withdrawal bleeding
occurred in response to a 10-day course of Provera (medroxyprogestrone
acetate, 10 mg/day). Partial dilatation and curettage initiated
shedding and contraction, and a biopsy revealed the uterine content to
be benign endometrial hyperplasia. This was attributed to unopposed
estrogen in the face of follicular phase levels of progesterone while
on RU486 (Table 2) (R. S. Newfield
et al., submitted for publication).
|
|
). After receiving
OCP therapy for 15 months, it was discontinued. Because her
menstruation was irregular for over a year after the discontinuation of
OCP, they were readministered. While receiving RU486, the patient experienced two other adverse effects that have been reported previously (13, 14, 15). The first was a transient erythematous rash that occurred 1 week after initiating therapy, which resolved while treatment continued. The second was mild Hashimoto’s thyroiditis occurring after months of therapy, with elevated antimicrosomal thyroid antibodies (1:1600 titer) and negative antithyroglobulin antibodies; T4 levels were low-normal (nadir of 4.5 ng/dL, normal range 4.5–11.0), with mildly elevated TSH levels (peak-7.3 mIU/mL, normal 0.4–4.0). Thyroid function tests returned to normal after withdrawal of RU486 therapy, but antimicrosomal thyroid antibodies remained elevated (as high as 1:6400 titer).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsCase HistoryResultsDiscussionConclusionReferences |
|---|
), these
ultimately proved to be intermittently dysregulated. At age 11.7 yr,
she exhibited a lack of response to CRF, which is probably not owed to
a basal cortisol level of 19 µg/dL or to poor quality of the ovine
CRF preparation, especially in view of the abnormal diurnal rhythm of
cortisol at that time. At age 16.5 yr the response to the CRF test
returned to normal. Nonetheless, even when the hormonal responses were
abnormal, levels of free cortisol in multiple 24-hr urine collections
were normal.
|
). Progesterone levels,
however, rose only slightly with RU486 treatment, but they remained in
follicular phase levels. Estradiol levels were unaltered. Despite
improvement of the patient’s striae, buffalo hump, and obesity in
response to RU486, BMD did not increase (Fig. 3).
GR sites for dexamethasone per cell and their fluctuations throughout
the course are presented in Table 1. In our studies, the normal number
of GR sites per cell for nonobese (n = 12) and obese (n = 7)
adolescent and young females was 5540 ± 878 and 11,078 ±
8057, with affinities (Kd) of 15.5 ± 6.7
nM and 33 ± 19.6 nM,
respectively. Over 2 yr, five determinations of the patient’s GR sites
per cell, before starting oral therapy with RU486 at the age of 13.8
yr, averaged 40,717 ± 16,191 (up to 63,241 sites/cell), with a
mean Kd of 24.8 ± 7.5 nM.
Analysis of these data by Scatchard plot showed evidence for a single,
high-affinity site. A therapeutic trial with RU486 was initiated at
high dose (400 mg/day) as an antiglucocorticoid agent. Results from
three different GR assays done before and after RU486 was introduced,
and the specific Scatchard plots from these assays, are shown in Fig. 4. In response to RU486, the purple
striae gradually resolved and cortisol levels rose, while GR sites per
cell dropped sharply to 5435 with a Kd of 36.2
nM. Treatment was then discontinued for 9 months
owing to persistent vaginal bleeding. Without RU486 treatment, GR
numbers remained normal, but striae soon reappeared. A few months
later, GR numbers gradually rose to 17,032. Following her second
vertebral fracture, RU486 was resumed and the striae again improved;
however, BMD (dual-energy x-ray absorptiometry) did not increase.
Although RU486 was discontinued 5 months later, owing to massive
endometrial hyperplasia (benign), the Cushingoid signs did not recur,
even though her GR rose from 10,787 while receiving RU486 to 33,979 1
week after RU486 was discontinued and to 23,893 8 weeks after that. The
lack of relapse within the first 1–2 weeks after withdrawal may be due
to a residual blockade by RU486, owing to its long half-life of 30–48
h (16). Indeed, 2 weeks after the first extended withdrawal (period E),
morning serum cortisol levels (33 µg/dL) were still elevated. Only at
3 weeks did they normalize. In period G, 1 week after the withdrawal of
RU486, the morning cortisol level was still high (59.6 µg/dL) and,
thus, precluded clinical relapse despite elevated GR (33, 979
sites/cell). However, by 8 weeks after RU486 had been withdrawn serum
cortisol was normalized (9.8 µg/dL). Cushingoid signs did nor recur
although GR were elevated (23, 893 sites/cell).
|
The GR promoter usage, established while the patient was off RU486, showed it to be the noncoding exon 1A, as in normal controls and in CEM-C7 cells. Even when our patient had normal GR site numbers/cell (period E), GR mRNA levels (145 pg/µg total RNA) were higher than in a control subject (90 pg/µg total RNA). In addition, there was not a linear correlation between GR site numbers and GR mRNA because a 2.4-fold rise in GR sites per cell (from 6,264 to 14,871) was only accompanied by a 1.2-fold rise in GR mRNA levels (from 145–175 pg/µg total RNA), suggesting a disordered posttranscriptional regulation of GR site numbers per cell.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsCase HistoryResultsDiscussionConclusionReferences |
|---|
). She did not manifest easy bruising or stunted
growth, two cardinal signs of Cushing’s syndrome in children. Evidence
for the latter is her growth pattern (Fig. 2), since commensurate with
a rapid weight gain starting at age 7 that preceded puberty, her height
velocity also increased. Her height predictions on presentation ranged
between 160 and 161 cm (according to the Bayley-Pinneau and
Roche-Wainer-Thissen methods) and are based on her bone age readings of
11 and 12.5 yr at chronological ages of 10.75 and 11.5 yr,
respectively. Her final height of 161 cm is below her midparental
target height of 167.5 cm, but it is only 2 cm below her father’s
corrected height. Furthermore, unlike the average 8.9 cm height deficit
(midparental–final height) in the group of 10 patients with Cushing’s
syndrome reported by Magiakou et al. (17), where vertebral
fractures were not described, a considerable part of the height deficit
in this patient must be attributed to the multiple vertebral fractures
she experienced. Therefore, we do not believe that her condition
affected her stature in precisely the same way as described by Magiakou
et al. (17).
She was normocortisolemic, yet on multiple occasions she demonstrated
no diurnal rhythm typical of cortisol (Table 3). Nonetheless, she had
normal urinary free cortisol levels. How can we explain this
paradox?
In the initial years of clinical investigation of this patient, before the initiation of RU486 treatment, there was correlation of the Cushingoid phenotype with a markedly increased number of GR sites per cell in peripheral lymphocytes (up to 63,241), but with normal binding affinity for dexamethasone. The affinity we determined is very close to that reported by several other groups for normal peripheral blood lymphocytes (reviewed in Ref. 18) and to what we have observed in normal and leukemic lymphoid cells using the whole cell assay. Peripheral Cushingoid signs in the presence of nondepressed cortisol levels might have been explained by differential GC sensitivity at the peripheral vs. the hypothalamic-pituitary level either by tissue-specific differences in GR numbers or by hypothalamic desensitization. In vitro evidence demonstrated that for cells with no inherent blocks in their response pathways, elevating the quantity of GR causes enhanced responses to glucocorticoids (20, 21). Such a scenario of differential GC sensitivity would resemble the rare cases of differential resistance to thyroid hormones. However, the transient nature of this condition and an HPA axis that was intermittently dysregulated do not support the interpretation that the sole cause of the clinical picture was elevated GR levels. Of note, RU486 does not alter the diurnal rhythm of cortisol (13) and, thus, did not contribute to HPA axis dysregulation. Furthermore, after RU486 treatment was initiated, the first extended withdrawal of therapy (period E) resulted in a prompt clinical relapse (e.g. striae) without an abrupt rise in GR numbers in the peripheral lymphocytes. In contrast, on final discontinuation of RU486 (period G), Cushingoid signs did not reappear despite a rise in GR that persisted for 2 months. Thus, GR numbers and phenotype no longer correlated. Could this inconsistency over time between GR and phenotype simply reflect the fact that RU486 altered the basic characteristics and natural history of this condition? This explanation is unlikely in view of the reappearance of Cushingoid signs as early as 3 weeks after RU486 withdrawal (period E). There was a correlation, however, between the number of GR sites per cell and the administration of RU486. Although RU486 generally resulted in a decrease in GR numbers, the clinical improvement (e.g. diminished and eventual resolution of striae) must be attributed to the antiglucocorticoid action of the drug. Although the HPA axis was disturbed in our patient, we did observe the expected HPA axis up-regulation after GR blockade with RU486; such an up-regulation does not occur in Cushing’s of nonpituitary origin (13, 14, 15, 21). The rise in cortisol might have been insufficient to counteract RU486 blockade in certain tissues, such that clinical improvement occurred (e.g. striae). However, the lack of improvement in BMD may again indicate variable tissue sensitivity, such that it is heightened at the level of the bone.
There was only one other case in the literature describing increased sensitivity to cortisol (22), which also remained a puzzle. It is similar in that the affected patient, a 54-year-old male with short stature (156.8 cm) had variable tissue sensitivity, presenting only with central obesity, moon face, and buffalo hump, yet no striae or reportable osteoporosis. He was also described having noninsulin-dependent diabetes mellitus since age 50. In addition, he had various abnormalities in either GR numbers (normal or elevated) or affinity (normal or decreased) that varied according to the ligand (cortisol vs. dexamethasone binding assays) and the cell type studied. The authors concluded that such abnormalities were not significant enough to unequivocally account for his condition. The cases are quite dissimilar in that the adult patient had an intact HPA axis and suppressed serum cortisol and urinary free cortisol levels and that his condition did not appear to be transient.
The transient nature of the Cushingoid features, and the fact that GR sequencing demonstrated that the primary GR amino acid sequence is normal, argues against a genetic defect at the receptor level. Indeed, although elevated GR numbers are clearly abnormal, they may not necessarily have caused the Cushingoid features in this patient. The explanation for the elevated GR noted initially remains unknown. In addition, our analysis showed normal promoter usage, thus appearing to rule out a switch to one of the other known promoter regions (which could contain unusual controlling sites).
One way to explain the puzzling characteristics of this case may be to view it as a maturational disorder that parallels the hormonal changes occurring during puberty, much like idiopathic juvenile osteoporosis, also a transient condition occurring between the ages of 8–14 yr (23). However, a postreceptor defect may be more plausible because it is now known that GR act in concert with other transcription factors, such as coactivators and corepressors. Although the 54-yr-old male with increased sensitivity to cortisol and an intact HPA axis discussed above had some abnormalities at the GR level, they were not judged to be primary. Rather, the authors speculated that a postreceptor defect might exist with the transcription machinery (22). Some transcription factors are also common to the GR as well as other steroid hormone receptors (24, 25). Therefore, it is possible that an abnormality in one of the proteins that constitute the transcription factor complex may explain our patient’s condition. For instance, the fluxes in sex steroids during puberty may temporarily alter the availability of transcription factors crucial for glucocorticoid action, and this derangement resolves once pubertal development is complete. A patient with a mild, preexisting imbalance between the various transcription factors may, thus, become transiently symptomatic only during a period of drastic hormonal changes, such as puberty.
|
Conclusion |
|---|
|
TopAbstractIntroductionMaterials and MethodsCase HistoryResultsDiscussionConclusionReferences |
|---|
|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 In remembrance of the excellent and most compassionate physician,
Dr. George Kalaitzoglou, who passed away in his prime years.
Received February 11, 1999.
Revised September 13, 1999.
Accepted September 28, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsCase HistoryResultsDiscussionConclusionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. Johanssen and B. Allolio Mifepristone (RU 486) in Cushing's syndrome Eur. J. Endocrinol., November 1, 2007; 157(5): 561 - 569. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. F.C. van Rossum and S. W.J. Lamberts Polymorphisms in the Glucocorticoid Receptor Gene and Their Associations with Metabolic Parameters and Body Composition Recent Prog. Horm. Res., January 1, 2004; 59(1): 333 - 357. [Abstract] [Full Text]
|
|
|
|
 |
|
 O. Heikinheimo, T. Raivio, H. Honkanen, S. Ranta, and O. A. Janne Termination of Pregnancy with Mifepristone and Prostaglandin Suppresses Transiently Circulating Glucocorticoid Bioactivity J. Clin. Endocrinol. Metab., January 1, 2003; 88(1): 323 - 326. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Raivio, J. J. Palvimo, S. Kannisto, R. Voutilainen, and O. A. Janne Transactivation Assay for Determination of Glucocorticoid Bioactivity in Human Serum J. Clin. Endocrinol. Metab., August 1, 2002; 87(8): 3740 - 3744. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. W. Tomlinson, N. Draper, J. Mackie, A. P. Johnson, G. Holder, P. Wood, and P. M. Stewart Absence of Cushingoid Phenotype in a Patient with Cushing's Disease due to Defective Cortisone to Cortisol Conversion J. Clin. Endocrinol. Metab., January 1, 2002; 87(1): 57 - 62. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Lacroix, N. N'Diaye, J. Tremblay, and P. Hamet Ectopic and Abnormal Hormone Receptors in Adrenal Cushing's Syndrome Endocr. Rev., February 1, 2001; 22(1): 75 - 110. [Abstract] [Full Text]
|
|
|
|
 |
|
 M. Boscaro, L. Barzon, and N. Sonino The Diagnosis of Cushing's Syndrome: Atypical Presentations and Laboratory Shortcomings Arch Intern Med, November 13, 2000; 160(20): 3045 - 3053. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
