Unrecognized adrenal insufficiency in patients undergoing laparoscopic adrenalectomy
Jamie Mitchell . German Barbosa . Michael Tsinberg . Mira Milas . Allan Siperstein · Eren Berber
Received: 16 May 2008/ Accepted: 1 October 2008/Published online: 27 November 2008 @ Springer Science+Business Media, LLC 2008
Abstract
Background Subclinical Cushing’s syndrome (SCS) is a well-described phenomenon where abnormalities of the hypothalamic-pituitary-adrenal axis exist in the absence of overt signs and symptoms of classic Cushing’s syndrome. While this has been shown to exist in 5-20% of patients with adrenal lesions, no standardized biochemical regimen exists to screen for SCS. Consequently, many of these patients may not be diagnosed prior to adrenalectomy with the risk of postoperative adrenal insufficiency. We began checking morning (a.m.) serum cortisol levels on post operative day 1 (POD1) following unilateral adrenalec- tomy for nonfunctioning adrenal lesions to determine the incidence of unrecognized adrenal insufficiency (AI) in these patients.
Methods One hundred and five patients undergoing adrenalectomy at a tertiary care center from 1999 to 2007 were retrospectively evaluated. Patients with Cushing’s syndrome, conditions associate with bilateral disease, and those receiving perioperative steroids were excluded, leaving 41 patients for analysis. A.m. serum cortisol levels were obtained in all patients POD1. Multiple factors were analyzed as possible predictors of AI. Analysis of variance (ANOVA), t-test, and chi-square test were used to deter- mine statistical significance.
Results The 41 patients’ diagnoses included 13 pheochro- mocytomas, 15 nonsecreting adenomas, 5 aldosteronomas, 5 metastatic lesions, 1 adrenocortical carcinoma, and 2 other benign lesions. Three groups were identified based on POD1, a.m. cortisol levels: sufficient (>10 µg/dl; n = 25, 61%), low-normal (3.4-10 µg/dl; n = 7, 17%), and insufficient (<3.4 µg/dl; n = 9, 22%). Tumor size and presence of diabetes, hypertension, and obesity were predictive of post- operative AI (p < 0.05).
Conclusions AI after unilateral adrenalectomy without evidence of cortisol hypersecretion on preoperative screening was present in a significant number of patients in our series. Patients with diabetes, hypertension, obesity, and larger tumors may be at higher risk for postoperative AI. More thorough screening for cortisol hypersecretion may be warranted in patients with these characteristics, and obtaining routine postoperative cortisol levels may avoid potentially dangerous unrecognized adrenal insufficiency following adrenalectomy.
Keywords Laparoscopy . Adrenalectomy . Insufficiency . Abdominal . Cushing’s syndrome . Subclinical
Incidentally discovered adrenal lesions, or adrenal “inci- dentalomas,” have become a much more common occurrence with the escalating use of high-quality imaging studies in clinical medicine. The majority of these lesions are benign, asymptomatic tumors which do not require any specific therapy beyond periodic monitoring. A fraction of these tumors are functional, resulting in hormonal excess and pathologic endocrine syndromes, such as pheo- chromocytoma (catecholamines), Conn’s syndrome (aldosterone) or Cushing’s syndrome (cortisol) [1].
J. Mitchell . G. Barbosa . M. Tsinberg . M. Milas .
A. Siperstein · E. Berber Endocrinology and Metabolism Institute, Section of Endocrine Surgery A-80, The Cleveland Clinic, 9500 Euclid Avenue, Cleveland, OH 44195, USA
e-mail: berbere@ccf.org
J. Mitchell
e-mail: mitchej5@ccf.org
Upon incidental detection of an adrenal lesion, a panel of screening laboratory tests is obtained to rule out hypersecretion of the aforementioned adrenal hormones. These include serum/urine metanephrines, plasma renin activity and aldosterone levels, and morning (a.m.) serum cortisol and 24-h urinary free cortisol levels [2]. While the screening tests for pheochromocytomas and aldosterono- mas have remained unchanged for many years, there have been recent reports which call into question the adequacy of a.m. serum cortisol and 24-h urinary free cortisol mea- surements in the screening of autonomous cortisol hypersecretion by adrenal cortical adenomas [3]. This is due to the recent characterization of the clinical entity termed subclinical Cushing’s syndrome (SCS), first described in 1981 by Charbonnel and colleagues [4]. SCS refers to presence of autonomous secretion of cortisol by an adrenal cortical tumor in the absence of the clinical man- ifestations of classic Cushing’s syndrome, such as moon facies, buffalo hump, proximal muscle weakness, and purple striae [5]. This has become much more recognized in recent years, with published reports documenting an incidence ranging from 5% to 20% of incidentally dis- covered adrenal lesions, and a prevalence of 79 cases per 100,000 persons [6, 7].
While serum a.m. cortisol and 24-h urinary free cortisol levels are felt to be adequate screening tests to rule out overt Cushing’s syndrome, they may not be sensitive enough to diagnose most patients with SCS [7, 8]. More sensitive testing has been suggested, such as the low-dose dexamethasone suppression test, in order to correctly diagnose these patients [8, 9]. While patients with SCS do not manifest obvious characteristics of classic Cushing’s syndrome, there are several important reasons for accu- rately diagnosing these patients. The first pertains to the surgical indications for resection of adrenal masses. Gen- erally, functional tumors and those greater than 4 cm in size warrant surgical removal [10]. While it is not entirely clear what the natural history of untreated SCS is, and it is believed that progression to classic Cushing’s syndrome is extremely rare [11, 12], there are several reports showing higher incidence of obesity, glucose intolerance, and arte- rial hypertension in patients with SCS [13-15]. This may argue for surgical resection in these patients when not otherwise indicated based on size or worrisome imaging characteristics [16].
The second important reason for diagnosing patients with SCS is that even subtle cortisol hypersecretion from an adrenal mass can lead to relative suppression of the contralateral gland, resulting in potentially life-threatening adrenal insufficiency following surgical resection. This has been described in several published reports on SCS [17, 18]. Knowledge of the presence of subtle cortisol
hypersecretion preoperatively allows appropriate steroid replacement therapy both during and after surgery.
Our clinical practice has been to screen all patients with adrenal masses using serum and urine metanephrines, serum a.m. and 24-h urinary free cortisol levels, and plasma aldosterone and renin activity levels. Several years ago, we began obtaining serum cortisol levels on the morning of postoperative day 1 in an effort to determine the incidence of adrenal insufficiency in patients under- going unilateral adrenalectomy. The aims of this study were to review our series of patients undergoing adrenal- ectomy to determine the incidence of adrenal insufficiency postoperatively in patients with normal preoperative serum and urine cortisol levels, and to determine the existence of any clinical predictors of this phenomenon.
Patients and methods
All patients undergoing adrenalectomy between 1999 and 2007 at the Cleveland Clinic by three endocrine surgeons were retrospectively reviewed. All patients were imaged preoperatively with computed tomography (CT) scans and had preoperative a.m. serum cortisol levels and 24-h urine collections for urinary free cortisol level measurements. Patients with Cushing’s syndrome were excluded from analysis. Additionally, patients with bilateral disease requiring bilateral adrenalectomy, hereditary syndromes associated with bilateral disease, such as multiple endo- crine neoplasia type 2 (MEN 2), von Hippel-Lindau syndrome (VHL), and neurofibromatosis type 1 (NF1), and those receiving steroids for indications unrelated to their adrenal disease were also excluded from further analysis.
All patients included in the study group had serum cor- tisol levels measured on the morning of postoperative day 1 to determine the incidence of adrenal insufficiency. Based on these levels, patients were divided into three groups: sufficient (group 1, >10 µg/dl), low-normal (group 2, 3.4- 10 µg/dl), and insufficient group 3, <3.4 µg/dl). All patients with adrenal insufficiency were placed on hydro- cortisone postoperatively and referred to an endocrinologist for gradual tapering of the dose. Multiple variables were analyzed with respect to preoperative biochemical profile, patient characteristics, and tumor characteristics in an attempt to identify differences between these groups and any factors which might be predictive of postoperative adrenal insufficiency in these patients. Statistical signifi- cance was determined using Student’s t-test, chi-square, Fischer’s exact test, and one-way analysis of variance (ANOVA), as appropriate. All data are expressed as mean ± standard error of the mean (SEM).
Results
Study group
During the study period 105 patients underwent adrenal- ectomy at our institution. Fourteen patients had Cushing’s syndrome and were excluded from further analysis. Five patients had bilateral adrenal disease (one VHL, one MEN 2, one NF, and two congenital adrenal hyperplasia) and were also excluded. Thirty-two patients did not have postoperative cortisol levels, and 13 were receiving ste- roids for reasons unrelated to their adrenal disease. Twelve of the 13 patients excluded for steroid use received pre- operative dexamethasone by the anesthesiology team for the treatment of postoperative nausea. This left 41 patients for analysis. In these patients, the pathologic diagnoses were as follows: 13 pheochromocytomas, 15 nonsecreting adrenal cortical adenomas, 5 aldosterone-secreting adeno- mas, 5 metastases from other primary tumors (1 esophageal squamous cell carcinoma, 1 hepatocellular carcinoma, 1 melanoma, 1 renal cell carcinoma, and 1 non-small-cell lung carcinoma), 1 adrenocortical carcinoma, and 2 other benign tumors. Pathologic diagnoses are summarized in Table 1. The mean age of the study population was 52.8 years, with an equal gender distribution (21 females, 20 males). Tumors occurred with equal frequency in the left (23, 56%) and right (18, 44%) adrenal glands. Mean length of hospital stay following surgery was 1.8 ± 0.2 days (range 1-8 days), with 66% of patients being discharged on postoperative day 1.
| Pathologic diagnosis | Number of lesions | Percentage of total |
|---|---|---|
| Pheochromocytoma | 13 | 32 |
| Nonsecreting adenoma | 15 | 37 |
| Aldosteronoma | 5 | 12 |
| Metastasis | 5 | 12 |
| Esophageal CA | (1) | |
| Hepatocellular CA | (1) | |
| Melanoma | (1) | |
| Renal cell CA | (1) | |
| Non-small-cell lung CA | (1) | |
| Adrenocortical carcinoma | 1 | 2 |
| Other benign lesions | 2 | 5 |
| Schwannoma | (1) | |
| Cystic lymphangioma | (1) | |
| Total | 41 | 100 |
CA, cancer
Patient characteristics
Cortisol levels
There were 25 patients with sufficient postoperative corti- sol levels (group 1, 61%), 7 patients with low-normal cortisol levels (group 2, 17%), and 9 patients with insuf- ficient levels (group 3, 22%). Postoperative cortisol levels were 15.2 ± 1.3 µg/dl for group 1, 8.2 ± 0.7 µg/dl for group 2, and 1.3 ± 0.4 µg/dl for group 3. When comparing preoperative a.m. serum cortisol levels between these groups, there were no significant differences, with mean values of 14.2 ± 2.6, 19.0 ± 4.2, and 11.6 ± 1.5 µg/dl, respectively (p = 0.14). Similarly, there were no signifi- cant differences in preoperative 24-h urine free cortisol levels between groups, with mean values of 43.9 ± 8.5, 59.6 ± 15.2, and 54.5 ± 20.1 ug/dl, respectively (p = 0.34). Additionally, there was no difference between the groups in preoperative adrenocorticotropic hormone (ACTH) levels (13.4, 11.0, and 10.0 pg/ml, respectively; p = 0.56). Patients in group 3 required steroid replacement postoperatively for an average of 4 months (range 2-7 months). Pre- and postoperative cortisol levels are sum- marized in Table 2.
Comorbidities
The patients with lower postoperative day 1 cortisol values had higher body mass index (BMI), with values of 27 ± 0.8 kg/m2 for group 1, 33 ± 1.8 kg/m2 for group 2, and 35 ± 3.2 kg/m2 for group 3 (p < 0.01). Incidence of diabetes also was found to increase as postoperative cor- tisol levels decreased, with 16% in group 1, 43% in group 2, and 78% in group 3 (p = 0.003).
When excluding patients with pheochromocytomas, there was a significantly higher incidence of arterial hypertension in group 3 (89%) than in the other two groups (43%, p = 0.024). Patient comorbidities with respect to postoperative cortisol levels are summarized in Table 3.
Age/gender
Although the patients with lower postoperative day 1 cor- tisol levels were younger, this difference was not statistically significant. There was no difference between the three groups with respect to gender.
Tumor characteristics
Pathologic diagnosis
Incidence of adrenal insufficiency with respect to tumor type was 31% for pheochromocytomas, 20% for
| Group | POD1 serum (3.4-26.9 µg/dl) | Preop. serum (3.4-26.9 µg/dl) | Preop. 24-h urine (<85 µg/g creatinine) |
|---|---|---|---|
| Sufficient (n = 25) | 15.2 ± 1.27 | 14.16 ± 2.63 | 43.86 ± 8.47 |
| Low-normal (n = 7) | 8.23 ± 0.66 | 19.02 ± 4.20 | 59.60 ± 15.2 |
| Insufficient (n = 9) | 1.31 ± 0.41 | 11.64 ± 1.52NS | 54.50 ± 20.1NS |
All values are expressed as mean ± SEM. Serum cortisol units are in ug/dl. 24-h urine cortisol levels are in ug/day. POD1, postoperative day 1; NS, not significant compared with sufficient and low-normal values
| POD 1 adrenal function | BMI (mean ± SEM), kg/m2 | Diabetes (% affected) | Hypertension (% affected) |
|---|---|---|---|
| Sufficient | 27 ± 0.8 | 16% | 44% |
| Low-normal | 33 ± 1.8 | 43% | 43% |
| Insufficient | 35 ± 2.5* | 78%* | 89%* |
* p < 0.01 versus sufficient; * p < 0.05 versus sufficient
| Tumor type | Total no. of cases | Cases of AI | Percentage |
|---|---|---|---|
| Pheochromocytoma | 13 | 4 | 31 |
| Nonsecreting adenoma | 15 | 2 | 13 |
| Adrenocortical carcinoma | 1 | 0 | 0 |
| Aldosteronoma | 5 | 2 | 40 |
| Metastases | 5 | 1 | 20 |
| Other benign lesions | 2 | 0 | 0 |
| Benign lesions | 35 | 6 | 17 |
| Malignant lesions | 6 | 2 | 33 |
| Primary adrenal lesions | 36 | 7 | 20 |
| Extra-adrenal lesions | 5 | 1 | 20 |
AI, adrenal insufficiency
aldosteronomas and metastatic lesions, and 13% for non- secreting adenomas. None of these differences reached statistical significance. This included comparisons of benign (17%) versus malignant lesions (33%) and primary adrenal (20%) versus extra-adrenal tumors (17%). These data are summarized in Table 4.
Tumor size
Lower postoperative day 1 cortisol levels were associated with increasing tumor size, with values of 3.1 ± 0.3 cm for group 1, 3.9 ± 0.7 cm for group 2, and 5.3 ± 1.0 cm for group 3 (p > 0.03).
| POD 1 adrenal function | Tumor size (cm) (mean ± SEM) | Hounsfield units (mean ± SEM) |
|---|---|---|
| Sufficient | 3.1 ±0.3 | 21.89 ± 5.04 |
| Low-normal | 3.9 ± 0.7 | 23.00 ± 11.9 |
| Insufficient | 5.3 ± 1.0* | 34.00 ± 5.64NS |
* p < 0.03 compared with sufficient; NS, not significant compared with sufficient and low-normal groups
Hounsfield units
There was no significant difference between the three groups regarding the Hounsfield unit density of the tumors as measured by computed tomography (CT) scan. Tumor characteristics of the three groups are summarized in Table 5.
Multivariate analysis
The independent variables that were found to be statisti- cally significant on univariate analysis (BMI, presence of diabetes and hypertension, and tumor size) were analyzed using multivariate analysis. This showed tumor size to be a significant predictor of postoperative adrenal insufficiency (p < 0.05). BMI approached but did not reach statistical significance (p = 0.06).
Discussion
It is becoming increasingly clear that subtle autonomous secretion of cortisol by adrenal tumors occurs much more frequently than previously recognized. This is supported by the present study, in which 22% of patients had evidence of adrenal insufficiency following adrenalectomy despite normal preoperative serum and urine cortisol levels, in accordance with previously published series [6, 7, 9]. This finding has several important implications. In regards to whether presence of SCS should be an indication for sur- gical resection, the data from this study supports other published series showing higher rates of obesity, diabetes,
and hypertension with their inherent risks among patients with SCS. Given the evidence that adrenalectomy in these patients results in improvement of these metabolic parameters [16], this argues for increased vigilance in screening patients with small (<4 cm) apparently non- functioning adrenal tumors for subtle autonomous cortisol secretion, as these patients may benefit from adrenalectomy in the absence of other indications for resection.
The next question is then: what is the best screening regimen for diagnosing SCS? No specific combination of tests of the hypothalamic-pituitary-adrenal (HPA) axis has been shown to be more sensitive than the others in detecting SCS. Different methods of testing as well as differences in criteria for cortisol suppression between studies have made it difficult to arrive at a consensus [19]. Clearly, as evidenced by our study population, a.m. serum and 24-h urine free cortisol levels lacked the sensitivity to identify these patients. The addition of routine dexameth- asone suppression tests to all patients may improve the detection of SCS, but studies have shown false negatives with this test as well [7]. Additionally, there are barriers to the routine use of these tests due to cost and patient com- pliance. Given the significantly higher incidence of obesity, hypertension, and diabetes in patients with SCS in our and other studies, it may be prudent to embark on more thor- ough screening of alterations of the HPA axis in patients possessing these comorbidities. Further studies may be required to determine the optimal method for diagnosing patients with SCS.
The second important point to be gleaned from these data is the importance of checking routine postoperative day 1, a.m. serum cortisol levels in patients undergoing adrenalectomy. While there are many studies looking at the incidence of SCS in fairly large series of patients with adrenal incidentalomas, very few of these look at the incidence of adrenal insufficiency following surgery. In those that do, very few patients are considered. Addition- ally, most studies exclude from analysis patients with lesions secreting hormones of any kind, such as aldosterone or catecholamines. Our series included all such patients in the analysis, as well as patients with metastatic lesions to the adrenal. Surprisingly, six of the nine patients with adrenal insufficiency postoperatively (67%) had hormon- ally active lesions (two aldosteronomas, four pheochromocytomas), and one had a metastatic lesion (hepatocellular carcinoma). While these results vary from those published by Dr. Shen and colleagues in 2006, whose review of 331 adrenalectomies did not find AI in 108 al- dosteronomas, 71 pheochromocytomas, and 29 adrenal metastases [20], our data suggests that it may be important to screen for postoperative AI in patients undergoing resection for any of these indications, not just those with apparently nonfunctioning primary adrenocortical lesions.
The finding of AI in patients with pheochromocytomas, aldosteronomas, and metastases begs the question of what mechanism explains the suppression of the contralateral adrenal gland. In the setting of adrenalectomy for metas- tases, it is feasible that the contralateral gland is affected by metastatic disease as well to such an extent as to interfere with its normal function, even if not detectable by con- ventional imaging studies such as CT or magnetic resonance imaging (MRI). Tumors with aldosterone hypersecretion may also have been subtly hypersecreting cortisol as well, with both hormones being produced in the adrenal cortex. Reports have been published of adrenal tumors co-secreting multiple adrenal hormones, although these usually are in the setting of adrenocortical cancer [21].
When considering a mechanism of contralateral adrenal suppression in the setting of a pheochromocytoma, this seems intuitively more difficult to explain given the fact that pheochromocytomas are medullary lesions whereas cortisol-secreting tumors are cortical lesions. However, reports have been published reporting evidence that pheo- chromocytomas can secrete corticotropin-releasing factor (CRF), which stimulates the secretion of ACTH. This could be a possible mechanism of increased cortisol secretion in the setting of a pheochromocytoma. Additionally, pheo- chromocytomas can secrete proteins called urocortins. These are CRF-related peptides which have adrenal receptors and can also result in increased cortisol secretion [22-24]. Cases of pheochromocytomas with concomitant cortisol hypersecretion have been described [22]. This could represent a possible explanation for the findings in the present study.
We employed the use of an a.m. serum cortisol level as a screening test for AI following adrenalectomy. The gold standard for making this diagnosis is the insulin tolerance test, in which patients are injected with insulin to induce hypoglycemia and serum cortisol and ACTH levels are drawn at 0, 20, 30, 45, 60, and 90 min. Due to its expense, cumbersome nature, and contraindications (cardiovascular disease, cerebrovascular disease, diabetes) this test has been replaced in clinical practice by the cosyntropin (ACTH) stimulation test. This test consists of injection of synthetic ACTH (1 or 250 µg) with serum cortisol levels obtained at 30 and 60 min. Failure of cortisol levels to increase appropriately is considered a positive result [25]. Only patients who were found to have AI postoperatively and subsequently required longer-term steroid replacement had an ACTH stimulation test. This decision was directed by the endocrinologists managing the steroid taper and occurred in two patients, one who required steroids for 16 weeks and one for 52 weeks, both of whom were operated on for pheochromocytomas.
While this test is considered more sensitive than a.m. serum cortisol levels, which represent the peak of
2 Springer
endogenous cortisol production, the latter is often used as a screening test with employment of the more sensitive test in the presence of ambiguity or high clinical suspicion in the absence of a frankly abnormal a.m. serum cortisol level. Additionally, several studies have compared the use of serum a.m. cortisol levels with the insulin tolerance test and found that the results correlated fairly well when the appropriate cutoff values were used for abnormal a.m. cortisol levels [26, 27]. Because of this, as well as the fact that a.m. serum cortisol levels are easier to obtain and less expensive, we used this as a screening test to diagnose AI postoperatively. It may be worthwhile to compare the results of this test with those of the ACTH stimulation test for the diagnosis of AI in patients undergoing adrenalec- tomy in the future.
Our data also showed that, in addition to a higher inci- dence of obesity, diabetes, and hypertension in patients with postoperative adrenal insufficiency, this group of patients tended to have larger tumors than patients with adequate adrenal function following adrenalectomy. This would not affect the decision to surgically remove these tumors, as they would already meet criteria for resection based on size. However, this suggests that patients with larger adrenal tumors deserve more thorough screening for SCS preoperatively, or more attention paid to the detection of adrenal insufficiency postoperatively.
In summary, this study shows that adrenal insufficiency following adrenalectomy may occur more frequently than previously realized. It is likely that these patients have undiagnosed subclinical Cushing’s syndrome, not detected on routine screening test employed preoperatively. SCS may be present in patients not only with apparently non- secreting adenomas, but those with adrenal metastases and hormonally functional cortical and medullary tumors, such as pheochromocytomas. Patients with characteristics such as large tumor size, or the presence of diabetes, obesity, and hypertension, may be at higher risk for postoperative AI and may benefit from more thorough screening mea- sures for SCS preoperatively. Additionally, due to the variability in sensitivity of various screening tests for SCS, it may be prudent to obtain postoperative day 1, a.m. serum cortisol levels in all patients undergoing adrenalectomy to identify potentially dangerous adrenal insufficiency.
References
1. Kuruba R, Gallagher SF (2008) Current management of adrenal tumors. Curr Opin Oncol 20:34-46
2. Petersenn S, Unger N, Walz MK, Mann K (2006) Diagnostic value of biochemical parameters in the differential diagnosis of an adrenal mass. Ann N Y Acad Sci 1073:348-357
3. Reincke M, Nieke J, Krestin GP et al (1992) Preclinical Cush- ing’s syndrome in adrenal “incidentalomas”: comparison with adrenal Cushing’s syndrome. J Clin Endocrinol Metab 75:826- 832
4. Charbonnel B, Chatal JF, Ozanne P (1981) Does the corticoad- renal adenoma with “pre-Cushing’s syndrome” exist? J Nucl Med 22:1059-1061
5. Sippel RS, Chen H (2004) Subclinical Cushing’s syndrome in adrenal incidentalomas. Surg Clin North Am 84:875-885
6. Reincke M (2000) Subclinical Cushing’s syndrome. Endocrinol Metab Clin North Am 29:43-56
7. Mantero F, Terzolo M, Arnaldi G et al (2000) A survey on adrenal incidentaloma in Italy. Study Group on Adrenal Tumors of the Italian Society of Endocrinology. J Clin Endocrinol Metab 85:637-644
8. Grumbach MM, Biller BM, Braunstein GD et al (2003) Manage- ment of the clinically inapparent adrenal mass (“incidentaloma”). Ann Intern Med 138:424-429
9. Mantero F, Masini AM, Opocher G et al (1997) Adrenal inci- dentaloma: an overview of hormonal data from the National Italian Study Group. Horm Res 47:284-289
10. Young WF Jr (2007) Clinical practice. The incidentally discov- ered adrenal mass. N Engl J Med 356:601-610
11. Ross NS (1994) Epidemiology of Cushing’s syndrome and sub- clinical disease. Endocrinol Metab Clin North Am 23:539-546
12. Terzolo M, Osella G, Ali A et al (1998) Subclinical Cushing’s syndrome in adrenal incidentaloma. Clin Endocrinol (Oxf) 48:89-97
13. Rossi R, Tauchmanova L, Luciano A et al (2000) Subclinical Cushing’s syndrome in patients with adrenal incidentaloma: clinical and biochemical features. J Clin Endocrinol Metab 85:1440-1448
14. Leibowitz G, Tsur A, Chayen SD et al (1996) Pre-clinical Cushing’s syndrome: an unexpected frequent cause of poor gly- caemic control in obese diabetic patients. Clin Endocrinol (Oxf) 44:717-722
15. Chiodini I, Torlontano M, Scillitani A et al (2005) Association of subclinical hypercortisolism with type 2 diabetes mellitus: a case- control study in hospitalized patients. Eur J Endocrinol 153:837- 844
16. Mitchell IC, Auchus RJ, Juneja K et al (2007) “Subclinical Cushing’s syndrome” is not subclinical: improvement after adrenalectomy in 9 patients. Surgery 142:900-905 discussion 905 e901
17. Emral R, Uysal AR, Asik M et al (2003) Prevalence of subclin- ical Cushing’s syndrome in 70 patients with adrenal incidentaloma: clinical, biochemical and surgical outcomes. Endocr J 50:399-408
18. Huiras CM, Pehling GB, Caplan RH (1989) Adrenal insufficiency after operative removal of apparently nonfunctioning adrenal adenomas. JAMA 261:894-898
19. Terzolo M, Bovio S, Pia A et al (2007) Subclinical Cushing’s syndrome. Arq Bras Endocrinol Metabol 51:1272-1279
20. Shen WT, Lee J, Kebebew E et al (2006) Selective use of steroid replacement after adrenalectomy: lessons from 331 consecutive cases. Arch Surg 141:771-774 discussion 774-776
21. Messer CK, Kirschenbaum A, New MI et al (2007) Concomitant secretion of glucocorticoid, androgens, and mineralocorticoid by an adrenocortical carcinoma: case report and review of literature. Endocr Pract 13:408-412
22. Kageyama K, Sakihara S, Yamashita M et al (2008) A case of multiple endocrine neoplasia type II accompanied by thyroid medullary carcinoma and pheochromocytomas expressing corti- cotropin-releasing factor and urocortins. Am J Med Sci 335:398- 402
23. Willenberg HS, Bornstein SR, Hiroi N et al (2000) Effects of a novel corticotropin-releasing-hormone receptor type I antagonist on human adrenal function. Mol Psychiatry 5:137-141
24. Fukuda T, Takahashi K, Suzuki T et al (2005) Urocortin 1, urocortin 3/stresscopin, and corticotropin-releasing factor recep- tors in human adrenal and its disorders. J Clin Endocrinol Metab 90:4671-4678
25. Arlt W (2008) Adrenal insufficiency. Clin Med 8:211-215
26. Hagg E, Asplund K, Lithner F (1987) Value of basal plasma cortisol assays in the assessment of pituitary-adrenal insuffi- ciency. Clin Endocrinol (Oxf) 26:221-226
27. Jones SL, Trainer PJ, Perry L et al (1994) An audit of the insulin tolerance test in adult subjects in an acute investigation unit over one year. Clin Endocrinol (Oxf) 41:123-128