| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
EXTENSIVE CLINICAL EXPERIENCE |
Departments of Endocrinology (Y.S.W., A.M.I., W.Z.W., G.A.K., P.J.J., J.P.M., G.M.B., A.B.G.) and Neurosurgery (F.A., I.S.), St. Bartholomew’s Hospital, London ECIA 7BE, United Kingdom
Address all correspondence and requests for reprints to: Prof. Ashley B. Grossman, Department of Endocrinology, St. Bartholomew’s Hospital, London EC1A 7BE, United Kingdom. E-mail: A.B.Grossman{at}qmul.ac.uk.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Objective: Our objective was to establish the clinical and biochemical characteristics of macroadenomas associated with Cushing’s disease compared with a large cohort of microadenomas and to assess their response to therapy.
Design: We conducted a retrospective case-records study for the years 1964–2001.
Setting: The study occurred at a tertiary referral hospital center.
Patients: Patients had Cushing’s disease presenting with a pituitary macroadenoma, in comparison with a large group of microadenoma patients.
Interventions: Interventions included therapy with surgery and radiotherapy.
Main Outcome Measures: Outcome measures included basal and dynamically responsive plasma ACTH and cortisol levels and response to treatment.
Results: We identified 18 patients with Cushing’s disease secondary to a macroadenoma; basal 0900 h plasma ACTH was 135.8 ± 32.5 and 45.0 ± 4.3 ng/liter (mean ± SEM), respectively, in macroadenomas and microadenomas (P = 0.013). Mean 0900 h serum cortisol was significantly increased in the macroadenomas (27.5 ± 3.0 µg/dl, 759.6 ± 82.6 nmol/liter, vs. 22.6 ± 0.6 µg/dl, 624.7 ± 16.4 nmol/liter) (P = 0.021). Testing with high-dose dexamethasone showed less suppression in the macroadenomas (57.6 ± 8.7% vs. 74.4 ± 2.1%; P = 0.02) and an attenuated ACTH response to CRH. For all biochemical variables there was considerable overlap between the two groups. Few patients with macroadenomas were cured by surgery.
Conclusions: Pituitary macroadenomas causing Cushing’s disease have biochemical features largely distinct from patients harboring microadenomas but represent one end of a continuum.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Few published series have focused specifically on macroadenomas, but such reports that exist have suggested that these patients show less suppressibility with high-dose dexamethasone, whether assessed by urinary free cortisol excretion (3) or plasma cortisol levels (1, 4). It has also been reported that such patients may have higher plasma ACTH levels than those with pituitary microadenomas (4, 5), but the data on cortisol levels, either urinary or circulating, have been inconsistent, as has the serum cortisol response to CRH stimulation (1, 3). However, the control group of microadenomas used as a comparison group has usually been limited in size. Furthermore, there is evidence that macroadenomas causing Cushing’s disease are more refractory to surgical treatment and show variable responsivity to adjunctive radiotherapy or medical therapy (1, 6).
The aim of this study was thus to compare the biochemical characteristics of pituitary macroadenomas causing Cushing’s disease compared with a large group of microadenomas and to assess the treatment outcome of macroadenomas. All the data were collected from St. Bartholomew’s Hospital, a tertiary referral center in London, over the years 1964–2001.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
A retrospective review of patients who were entered in the Cushing’s syndrome database was performed, with the authorization of our institutional case-note review committee, and a total of 413 patients with a diagnosis of Cushing’s syndrome at St. Bartholomew’s Hospital in the years 1964–2001 were identified. Of the 413 patients, 274 patients had ACTH-dependent Cushing’s syndrome of pituitary origin. However, the great majority of these patients have been identified over the past 20 yr, from 1985 onward. An analysis of the responses to CRH and the high-dose dexamethasone suppression tests (HDDST) in the group as a whole have been included in previous publications (7, 8).
Cushing’s disease was diagnosed on the basis of clinical symptoms and signs, an elevated midnight sleeping serum cortisol, and a failure to suppress serum cortisol to less than 1.8 µg/dl (50 nmol/liter) on a low-dose dexamethasone suppression test. The source of excessive cortisol production was localized using imaging studies and, where appropriate, bilateral inferior petrosal sinus sampling in the more recent patients. ACTH-secreting macroadenomas were defined on either computed tomography (CT) scanning or magnetic resonance imaging (MRI) as showing a pituitary tumor greater than 1 cm in at least one dimension, with subsequent confirmation of the pathological diagnosis of a pituitary adenoma immunostaining for ACTH. Details of the scanning were noted, but because of different machines over the course of the study, a direct comparison of tumor size could not be made.
We elected to include all patients with microadenomas and Cushing’s disease from the total database as our control group for comparison. The diagnosis of a pituitary microadenoma was based on the previously mentioned criteria for Cushing’s disease plus the absence of a macroadenoma on imaging; this therefore included all patients with biochemical evidence of ACTH-dependent Cushing’s syndrome with radiological evidence of a pituitary tumor less than 1 cm (including patients with no obvious tumor) and either pathological evidence of pituitary adenoma immunostaining for ACTH or a positive clinical and biochemical response to transsphenoidal surgery (TSS).
Methods
All patients underwent a complete biochemical and endocrine assessment using a standard protocol (9). Plasma cortisol was measured using a fluorometric method from 1977–1982 and then by unextracted nonchromatographic RIA from 1982–2000, with a working range set at 1.8–72.5 µg/dl (50–2000 nmol/liter). The coefficient of variation at 3.6 µg/dl (100 nmol/liter) and 36.2 µg/dl (1000 nmol/liter) was 6%. The fluorometric assay proved to be equivalent to subsequent assays down to 5.4 µg/dl (150 nmol/liter), but below this level there was nonspecific fluorescence. From 2000 onward, serum cortisol was measured by competitive immunoassay format on the fully automated Technicon Immuno-1 analyzer (Bayer, Newbury, UK). The lower limit was set at less than 1.8 µg/dl (50 nmol/liter), with an interassay coefficient of variation of less than 3.5%. The two assays after 1982 were validated against each other using control sera and found to be equivalent. Only a small number of patients were assessed by the assay before 1982. Before 2002, for ACTH we had used an in-house extracted RIA, using ACTH-(1–39) reference standards from Bachem Chemicals (Torrance, CA); this showed substantial cross-reaction with proopiomelanocortin (POMC) but not with ACTH fragments (10). The inter- and intraassay coefficients of variation were less than 8%.
The dexamethasone suppression test employed the use of oral administration of 0.5 mg (low dose) or 2 mg (high dose) dexamethasone every 6 h for 2 d (9). For the CRH stimulation test, 100 µg human-sequence CRH was administered iv. In this study, we applied the CRH test response criteria reported in our previous publication, which is a rise in cortisol above a mean basal level of 20% (8).
We identified all patients from the total group of patients with Cushing’s disease who had imaging abnormalities consistent with a macroadenoma and assessed their charts for detailed evaluation of biochemical diagnostic tests and of their treatment outcome. The patients were regarded as cured postoperatively if 0900 h serum cortisol (within 2 wk after surgery) was undetectable, less than 1.8 µg/dl (<50 nmol/liter), on more than one occasion. The overall outcome of therapy was classified as 1) remission if there was hypoadrenalism or normal adrenal function and regression of stigmata of hypercortisolism, 2) controlled if normal adrenal function (defined as a mean cortisol on the basis of five values throughout the day in the range 5.4–10.9 µg/dl (150–300 nmol/liter) was maintained by medical therapy or had been obtained by bilateral adrenalectomy, and 3) death. The clinical data, biochemical parameters, and imaging characteristics of the tumors (e.g. suprasellar extension and cavernous sinus invasion) were compared between the cured and noncured patients with macroadenomas for identification of possible predictive factors.
Statistical analysis
The statistical analysis was performed using SPSS (version 11.5). The results are presented as means ± SEM. Because the data were normally distributed, the differences between macroadenomas and microadenomas were analyzed using the independent Student’s t test, 
2 test or Fisher’s exact test where appropriate. Stepwise multiple logistic regression was used to identify factors predictive of surgical outcome and remission.
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Clinical features of patients with a macroadenoma
Patients with macroadenomas had a mean age of 47.3 ± 13.4 yr (mean ± SD); 77.8% were female. At presentation, the major clinical features noted were moon face (94.4%), weight gain (77.8%), central obesity (72.2%), hypertension (72.2%), plethora (61.1%), bruising (61.1%), skin thinning (61.1%), and muscle weakness (61.1%); none of these features, including the mean age, was significantly different from those patients with microadenomas. In 27.8%, there was evidence of a visual field defect, but clinical pigmentation was uncommon (5.6%). Imaging revealed a suprasellar extension in eight of 18 (44%) and cavernous sinus invasion in seven of 18 (39%); in four patients, two with suprasellar extensions and two without, the tumor was at least partly cystic (Table 1
).
|
Serum cortisol levels at 0900 h were available in all patients with macroadenomas and microadenomas. For plasma ACTH, all patients had at least one value that was clearly detectable (>10 ng/liter). However, for statistical comparison, we decided to use plasma ACTH taken under strictly standardized conditions, and we therefore chose the basal plasma ACTH taken as the baseline value for the CRH test, which was taken at 0900 h, supine and fasting; such ACTH levels were available in all 18 patients with macroadenomas and in 68 patients with microadenomas. The serum cortisol and plasma ACTH levels were both significantly higher in the macroadenoma group than those with microadenomas (P = 0.021 and P = 0.013, respectively) (Table 2 and Figs. 1 and 2). In addition, the cortisol to ACTH ratio also showed a significant difference between the two groups, with the microadenomas showing relatively higher levels of cortisol for a given ACTH (Table 2). The full results of the HDDST were available in 17 patients with macroadenomas and 174 with microadenomas; the degree of cortisol suppressibility was significantly different between the two groups (P = 0.02) (Table 3), with the macroadenoma patients showing a lesser degree of suppression (57.6 ± 8.7%) compared with the microadenomas (74.4 ± 2.1%) (Fig. 3). A threshold of 50% or more suppression during the HDDST was seen in 11 of 17 (64.7%) patients with macroadenomas and 152 of 174 (87.4%) microadenoma patients (P = 0.023) (Table 3). The cortisol response to CRH stimulation in 17 patients with macroadenomas was compared with that of 73 patients with microadenomas; the percentage rise in serum cortisol above basal was less in the macroadenoma group compared with the microadenoma group, and fewer patients showed a response (defined as a mean rise above basal of >20%: 64.7% vs. 83.6%) (Table 4), but the difference did not attain statistical significance (P = 0.097). When the percentage ACTH response in the macroadenoma patients was compared with those in the microadenoma group (n = 68), the ACTH response to CRH was significantly attenuated in the macroadenoma patients (P = 0.03). Furthermore, the absolute levels of ACTH were higher at each time point in the macroadenoma patients [e.g. at 30 min, 230 ± 32 ng/liter (range, 54–447 ng/liter)] compared with 118 ± 12 ng/liter (range, 15–468 ng/liter) in the microadenomas, although with considerable overlap. Data for bilateral inferior petrosal sinus sampling were available in 11 patients with macroadenomas and 72 patients with microadenomas; whereas both basal and stimulated central ACTH levels were higher in patients with macroadenomas, this also failed to reach statistical significance (P = 0.28 and 0.083, respectively) (Table 5).
|
|
|
|
|
|
|
All patients with Cushing’s disease underwent TSS as first-line treatment. However, only two of 16 (12.5%) patients with macroadenomas were cured after initial TSS, and the remaining 14 not cured by TSS were referred for radiotherapy. In addition, they received adjunctive therapies including repeat surgery, conventional external-beam radiotherapy, medical therapy (metyrapone, ketoconazole, or mitotane depending on patient response and tolerance) and bilateral adrenalectomy, as appropriate (Fig. 4). In summary, five of 14 (35.7%) patients with macroadenomas went into remission with TSS and radiotherapy. We analyzed various factors including age, duration of symptoms, the presence of cavernous sinus invasion, suprasellar extension, serum 0900 h cortisol, plasma 0900 h ACTH, and the responses to the HDDST and CRH stimulation, but none was a significant predictor of postoperative cure or the likelihood of remission.
|
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
The baseline hormonal assessment of patients with macroadenomas showed a clear difference from the microadenomas, with both 0900 h ACTH and cortisol levels being significantly higher than those in microadenomas. Previous studies had indicated significantly higher ACTH levels in macroadenomas, but these studies had not shown any significant difference in urinary free cortisol or serum cortisol levels or in cortisol metabolites (3, 4). A linear correlation between maximum tumor diameter and baseline plasma ACTH level was demonstrated by Losa et al. (12), and a similar correlation between mean ACTH and tumor volume was reported by Selvais et al. (4). This has led to the speculation that the forms of ACTH present in macroadenomas may include precursors that are immunoreactive but not biologically active (4, 13), and this could be possible; ACTH as measured by the newer immunometric assays shows very little cross-reactivity with POMC, but all of our patients were assessed with an immunoassay which does indeed measure ACTH precursors in addition to ACTH-(1–39) (10). In both of the major series, there was a trend to increases in cortisol in patients with macroadenomas, and it is more likely that the relatively small sample sizes were not powered to produce significance; in our series, we indeed found statistically higher levels of cortisol in the patients with macroadenomas. Nevertheless, the increase in ACTH is more apparent in the macroadenoma patients, and indeed the ratio of ACTH to cortisol suggests that for a given level of ACTH, less cortisol is found in the macroadenoma patients. It is possible that this reflects lowered biological activity of ACTH in patients with macroadenomas, because the ACTH assay used cross-reacts with the ACTH precursor POMC. However, from the point of view of clinical practice, there is considerable overlap with the levels seen in microadenomas, and basal levels alone of either ACTH or cortisol are unable to reliably differentiate macroadenomas from microadenomas (Figs. 1
and 2).
The HDSST has been used to distinguish a pituitary from an ectopic source of ACTH with a sensitivity of 60–80% and high specificity when a cutoff in plasma cortisol suppression of more than 50% is used (14, 15), although we and others have argued that it no longer has much diagnostic utility in the differential diagnosis of the source of ACTH in ACTH-dependent Cushing’s syndrome (7, 16). Nevertheless, it has been used extensively in the past, and our data are concordant with those of others in demonstrating less glucocorticoid suppressibility in patients with macroadenomas compared with microadenomas (3, 4). Again, there is still considerable overlap with the microadenoma group (Fig. 3), suggesting that for an individual patient the test is of little use in differentiating a pituitary macroadenoma from a microadenoma. We and others have previously indicated that the HDDST is also an unreliable test in differentiating ectopic ACTH from Cushing’s disease. The responsivity to the CRH test is also compatible with lesser responses in the macroadenoma group; this was significant for the ACTH increment but not for serum cortisol. In part, this smaller rise in ACTH may represent a lesser response from an elevated baseline level. A failure to respond to CRH in macroadenoma patients had been reported in five patients with macroadenomas by Ikeda et al. (5)
These results suggest that, in Cushing’s disease, a macroadenoma is more likely to be associated with higher ACTH and cortisol levels and lesser responsivity to dynamic stimuli, indicating a more autonomous mode of secretion of ACTH, but there was considerable overlap with results from the microadenomas. These results may offer some insight into the basic molecular derangements in the disorder. Studies by Biller et al. (17) and Gicquel et al. (18) have shown that Cushing’s disease caused by both microadenomas and macroadenomas is monoclonal in origin; such tumors are considered to result from sporadic mutation and subsequent clonal proliferation of neoplastic corticotroph cells. Katznelson et al. (3) investigated whether there was enhanced proliferative potential in ACTH-secreting macroadenomas in comparison with microadenomas by assaying in situ for the presence of Ki-67, a cell cycle antigen and a marker of cell proliferation. They found a trend toward a higher degree of proliferative activity in macroadenomas. In addition, the degree of Ki-67 staining was shown to correlate with aggressiveness of the neoplasm; those tumors having high Ki-67 staining were found to have more invasive proliferative activity, whereas others have correlated Ki-67 staining with maximum tumor diameter (12). What remains unclear is whether these macroadenomas are simply one extreme of a spectrum of corticosteroid resistance or represent a separate type of tumor with specific growth and biochemical characteristics. Previous studies have suggested that the relative dexamethasone resistance characteristic of corticotroph tumors is unrelated to abnormalities in sequence or expression of the glucocorticoid receptor (19, 20). However, it has recently been shown that corticotrophs express mRNA for the ACTH receptor, and this may be involved in feedback regulation (21). In patients with Cushing’s disease, this receptor may be either lost or expressed to a lesser extent, especially in the presence of high ACTH levels (21). There was no clear difference between macroadenomas and microadenomas, but because tumors available for study will be of necessity the larger tumors, it is still possible that macroadenomas are the most likely to lose ACTH autoregulation and demonstrate the greatest autonomy; currently, however, this remains largely speculative.
Only two of 16 (12.5%) patients in the macroadenoma group were cured after the initial surgery. This is based upon using a strict criterion of very low serum cortisol level at 0900 h, 1.8 µg/dl (<50 nmol/liter) taken within 2 wk after surgery, because late relapses are rarely seen using this criterion (22). Cannavo et al. (1), defining surgical cure as transient or permanent hypoadrenalism after surgery and no tumor remnant found at last MRI, reported a cure rate of eight of 26 (30.8%) patients with macroadenomas. Blevins et al. (6), defining surgical cure as normalization of postoperative cortisol levels (postoperative serum cortisol < 5 µg/dl or urinary free cortisol < 15 µg/24 h), reported a 67% cure rate. The remission rate of Cushing’s disease after transsphenoidal surgery, without noting the size of the tumor but mostly including microadenomas, ranges from 76–92% in published reports (22, 23, 24, 25, 26, 27, 28, 29). The variability in the definition of cure inevitably makes direct comparison difficult, but our own previously published surgical cure rate of 52% in Cushing’s disease, using the same stringent criterion as used here (22), suggests that ACTH-secreting macroadenomas are indeed more refractory to surgical treatment. We also noted two perioperative deaths in this series of macroadenomas, which is a relatively high perioperative mortality rate.
The published remission rates after radiotherapy range from 23–83% for patients with Cushing’s disease, without taking into consideration the size of the tumor, but this may take a variable number of years to become effective (30, 31, 32). In the series from Blevins and colleagues (6), only one of eight (12.5%) patients remained in remission after radiotherapy, with the single patient in remission being followed up for only 9 months. In another study, radiotherapy was observed to be effective in 56% of patients with macroadenomas (1). In comparison, our study showed that 28.6% of patients with residual disease in the macroadenoma group remained in remission after radiotherapy with a mean follow-up period of 133 months. In addition, 14.2% of patients required medication to control hypercortisolism after radiotherapy at a mean follow-up of 61 months. However, this relatively low rate of hormonal control by radiotherapy may be contrasted with tumor stabilization, with only two of 14 patients subsequently showing signs of tumor growth after radiotherapy. It seems likely that although radiotherapy is rarely able to control hormonal hypersecretion effectively in the short term, it may be effective in the prevention of tumor progression. It may also be noted that four patients in our series also underwent bilateral adrenalectomy in addition to radiotherapy, and only one showed evidence of tumor progression (subsequently leading to her demise); the three surviving patients had been followed up for 59.7 months. Although the numbers are small, it is probable that the previous irradiation played a role in preventing tumor growth in these patients (31). Repeated surgical intervention for primary surgical failure has been shown to be ineffective in achieving permanent normalization of ACTH secretion (6, 33).
Different investigators have looked into which factors might predict treatment outcome (1, 6, 24, 34, 35). Bochicchio et al. (24) reported the multicenter results of transsphenoidal surgery of 668 patients and showed that tumor size, invasiveness, and extrasellar growth did not influence the outcome of surgery. However, they did not look at macroadenomas as a separate group. Other studies have reported that patients with macroadenomas showing cavernous sinus invasion and/or tumor size more than 2 cm correlated with an unfavorable outcome after surgery (1, 6). Cannavo et al. (1) also demonstrated that baseline ACTH levels served as a good predictive index for surgical outcome, with cured patients having significantly lower levels. They also found that the size of the tumor correlated closely with the level of ACTH. In this series, we failed to identify any factors predictive of surgical cure or remission, although it may be worth pointing out that neither of the two surgically cured patients had cavernous sinus invasion.
In conclusion, we have surveyed the clinical and diagnostic features of patients with Cushing’s disease and macroadenomas and report that such patients show significantly higher ACTH and cortisol levels and relative resistance to dexamethasone compared with patients with microadenomas, a proportionally smaller ACTH response to CRH, and a higher ratio of ACTH to cortisol. The surgical cure rate is also low, and whereas radiotherapy appears to prevent tumor progression, it infrequently led to hormonal remission. We therefore confirm that these tumors have a distinct profile compared with microadenomas, although they probably represent one end of a spectrum of tumor autonomy, and there is considerable overlap with the microadenoma group.
|
Footnotes |
|---|
Abbreviations: CT, Computed tomography; HDDST, high-dose dexamethasone suppression test; MRI, magnetic resonance imaging; POMC, proopiomelanocortin; TSS, transsphenoidal surgery.
Received January 12, 2005.
Accepted May 2, 2005.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  B. M. K. Biller, A. B. Grossman, P. M. Stewart, S. Melmed, X. Bertagna, J. Bertherat, M. Buchfelder, A. Colao, A. R. Hermus, L. J. Hofland, et al. Treatment of Adrenocorticotropin-Dependent Cushing's Syndrome: A Consensus Statement J. Clin. Endocrinol. Metab., July 1, 2008; 93(7): 2454 - 2462. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. F. Moore and S. K. Grinspoon A Dural Tale J. Clin. Endocrinol. Metab., September 1, 2007; 92(9): 3367 - 3368. [Full Text] [PDF]
|
|
|
|
|
|
A. Munir, F. Song, P. Ince, S. J. Walters, R. Ross, and J. Newell-Price Ineffectiveness of Rosiglitazone Therapy in Nelson's Syndrome J. Clin. Endocrinol. Metab., May 1, 2007; 92(5): 1758 - 1763. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M. Isidori, G. A. Kaltsas, C. Pozza, V. Frajese, J. Newell-Price, R. H. Reznek, P. J. Jenkins, J. P. Monson, A. B. Grossman, and G. M. Besser The Ectopic Adrenocorticotropin Syndrome: Clinical Features, Diagnosis, Management, and Long-Term Follow-Up J. Clin. Endocrinol. Metab., February 1, 2006; 91(2): 371 - 377. [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 |
