CLINICAL OUTCOMES IN ADRENAL INCIDENTALOMA: EXPERIENCE FROM ONE CENTER
Jekaterina Patrova, MD1,2,3; Iwona Jarocka, MD1,4; Hans Wahrenberg, PhD, MD1,5; Henrik Falhammar, PhD, MD1,2
ABSTRACT
Objective: To investigate the outcome in patients with adrenal incidentaloma (AI).
Methods: A retrospective evaluation of 637 patients with AI referred to a tertiary center over 8 years. Radiologic and hormonal evaluations were performed at baseline. Follow-up imaging was carried out if necessary, and hor- monal evaluation was performed at 24 months according to national guidelines.
Results: The mean age was 62.7 ± 11.6 years, and the mean AI size was 25.3 ± 17.0 mm at presentation. Hormonal evaluation revealed that 85.4% of all tumors were nonfunctioning adenomas, 4.1% subclinical Cushing syndrome (SCS), 1.4% pheochromocytoma, 1.4% primary hyperaldosteronism, 0.8% Cushing syndrome, 0.6% adre- nocortical carcinoma, 0.3% congenital adrenal hyperpla- sia, 2.2% metastasis to adrenals, and 3.8% other lesions of benign origin. Bilateral tumors were found in 11%, and compared to unilateral tumors, SCS was more prevalent. Only 2 cases were reclassified during follow-up, both as SCS, but neither had had a dexamethasone suppression test
From the 1Department of Endocrinology, Metabolism and Diabetes, Karolinska University Hospital, Stockholm, Sweden, 2Department of Molecular Medicine and Surgery, Karolinska Institutet, Stockholm, Sweden, 3Department of Endocrinology, Vilnius University Santariskes Klinikos, Vilnius, Lithuania, 4Department of Medicine, District General Hospital, Nyköping, Sweden, and 5Department of Medicine Huddinge, Karolinska Institutet, Stockholm, Sweden.
Address correspondence to Dr. Jekaterina Patrova, Department of Endocrinology, Vilnius University Santariskes Klinikos, Santariskiu 2 Vilnius LT-08661, Lithuania.
E-mail: Jekaterina.patrova@gmail.com.
Published as a Rapid Electronic Article in Press at http://www.endocrine practice.org on June 29, 2015. DOI: 10.4158/EP15618.OR
To purchase reprints of this article, please visit: www.aace.com/reprints. Copyright 2015 AACE.
performed at initial work-up. In patients diagnosed with an adrenal metastasis, 92.9% were deceased within 2 years. Excluding those with malignant tumors, 12.9% of patients died during the study period of up to 11 years due to other causes than adrenal.
Conclusion: Most AIs were benign, but a small frac- tion of tumors were functional and malignant. The prog- nosis of patients with adrenal metastasis was extremely poor, but otherwise, the mortality rate was similar to that for the general population. Follow-up of AIs <4 cm with an initial nonfunctional profile and benign radiologic appear- ance appears unwarranted, but screening for congenital adrenal hyperplasia should be considered. (Endocr Pract. 2015;21:870-877)
Abbreviations:
17OHP = 17-hydroxyprogesterone; ACC = adrenal cortical carcinoma; ACTH = adrenocorticotropic hor- mone; AI = adrenal incidentaloma; CAH = congenital adrenal hyperplasia; CT = computed tomography; CS = Cushing syndrome; DST = dexamethasone suppres- sion test; HPA = hypothalamic-pituitary-adrenal axis; MRI = magnetic resonance imaging; SCS = subclinical Cushing syndrome; T2DM = type 2 diabetes mellitus; UFC = urinary free cortisol
INTRODUCTION
The term adrenal incidentaloma (AI) was coined in 1982 (1) and was initially used to describe an incidentally discovered adrenal mass on imaging study ordered for con- ditions unrelated to any suspicion of adrenal disease. The development of and increased use of imaging techniques such as computed tomography (CT) and magnetic reso- nance imaging (MRI) has resulted in a sharp rise in the fre- quency of discovered AIs. AIs are rarely (<1%) detected in patients <30 years of age but are common (7%) in patients >70 years of age, and they are imposing an increasing burden on the health care system (1). Only malignant or
hyperfunctional AIs (excess production of cortisol, aldo- sterone, catecholamines, or adrenal androgens) need active treatment; thus, it is mandatory to identify these patients. A number of clinical series regarding AI have been reported in the literature; however, the most effective diagnostic procedures to distinguish between the various etiologies of AI have not been established (1-4). Thus, there is no international consensus regarding the follow-up and man- agement of AI. Follow-up for several years has been fre- quently advocated, but this has recently been challenged (3).
The primary aims of this study were to determine the prevalence of malignancy and hormonal activity among AIs, the prevalence of metabolic disturbances in AI-affected individuals with borderline cortisol hypersecretion (sub- clinical Cushing syndrome [SCS]), tumor growth and hor- mone secretion during follow-up, and mortality.
METHODS
This retrospective study was conducted at the Department of Endocrinology, Metabolism and Diabetes, Karolinska University Hospital, Stockholm, Sweden. Between January 1, 2003, and December 31, 2010, 767 patients had been referred with the diagnosis of AI. A total
of 130 cases were excluded because the tumors had been found during work-up for a suspected adrenal tumor or during the staging and follow-up of a known malignancy. However, cases in which adrenal metastasis was the ini- tial discovery of a malignancy were included. Hence, 637 patients fulfilled the criteria of a true AI and were included in the study (Fig. 1). All referred subjects were examined according to the then-current Swedish National Guidelines for Evaluating and Treating AI (4).
According to the National Guidelines, all patients underwent radiologic examination of the adrenal glands with a CT or MRI scan. Unenhanced and enhanced CT images were evaluated. A density ≤10 Hounsfield units (HU) on an unenhanced CT image was considered diag- nostic of lipid-rich adrenal adenoma. If the density was >10 HU, enhanced CT images were evaluated, and the washout after 15 minutes was calculated. An absolute percentage washout of ≥60% at 15 minutes was considered indicative of a benign cortical adenoma. The largest transverse diam- eter in the axial plane in a three-dimensional planar system was used to represent the lesion size.
Biochemical evaluation consisted of analysis of: serum cortisol at 8:00 AM, followed by overnight dexamethasone suppression test (DST) (1 mg given at 11:00 PM) and mea- surement of cortisol levels at 8:00 AM the next morning
Assessed for eligibility (n = 767)
Excluded due to not fulfilling the criteria for AI (n = 130)
Fulfilled the criteria for true AI
Biochemical evaluation performed:
P-Cortisol
NFA: n = 544 (85.4%, 63%
DST or 24-h UFC
P-Metanephrines or U-Catecholamines
Other: n = 24 (3.8%, 71% F)
Aldosterone/renin ratio
F)
Age: 58.0 ±16.2
Age: 62.9 ±11.4
Other tests to the discretion of the physician
Size: 39.5 ± 25.4
Size: 22.9 ±14.2
SCS: n = 26 (4.1%, 80% F)
CS:
PA:
Metastasis: n = 14 (2.2%, 57% F)
Pheo: n = 9 (1.4%, 56% F)
ACC: n = 4 (0.6%, 0% F)
CAH:
n = 5 (0.8%, 80% F)
n = 9 (1.4%, 33% F)
n = 2 (0.3%, 100% F)
Age: 61.6 ± 11.05
Age: 53.8 ±11.4 Size: 30.0 ±14.8
Age: 60.3 ±9.9
Age: 68.6 ± 10.0
Age: 64.9 ±8.7
Age: 64.0 ±4.7 Size: 91.2 ±34.2
Age: 32.5 ±24.7
Size: 35.6 ±16.3
Size: 23.9 ±13.5
Size: 39.1 ±23.0
Size: 44.6 ±21.8
Size: 32.5 ± 24.7
Size: The largest tumor diameter in millimeters
Age: the age of the patient in years
Other: Myelolipoma, cyst, adrenal bleeding
(<1.8 ug/dL was considered normal) or 24-hour urinary free cortisol (UFC); plasma metanephrines or 24-hour samples of urinary catecholamines; and plasma renin and aldosterone, with the aldosterone to renin ratio calculated for those patients with high blood pressure and/or hypo- kalemia. There was a shift over time from UFC analysis to DST as the primary screening test for hypercortisolism. Fasting plasma glucose (to detect impaired fasting glucose and diabetes) and blood pressure were also measured. At the discretion of the physician, additional biochemical analyses or examinations could be performed. Measuring of morning serum 17-hydroxyprogesterone (17OHP) level was only performed when one of the coauthors (H.F.) was screening the referrals, and then it was only done randomly.
Surgery was generally recommended if the AI was hormonally active or if >4 cm in size. If <4 cm, follow- up and treatment were planned according to the individual radiologic characteristics of the lesion. For lesions in which malignancy could not be ruled out by radiology, a CT scan was repeated after 3, 6, and 12 months. Further investi- gation of patients with suspicion of SCS was planned individually according to the severity of accompanying diseases (diabetes mellitus, hypertension, osteoporosis, obesity). SCS was defined as cortisol ≥1.8 ug/dL after DST in combination with at least one other abnormal test of the hypothalamic-pituitary-adrenal (HPA) axis. The second hormonal test confirming abnormal HPA function included plasma adrenocorticotropic hormone (ACTH) level <9 pg/mL at 8:00 AM, elevated midnight cortisol/patho- logic diurnal rhythm, or elevated UFC levels.
Cushing syndrome (CS) was defined as at least 2 first- line biochemical tests (DST, UFC, midnight serum cortisol) returning clearly abnormal results, together with a clinical picture of CS. A midnight serum cortisol <1.8 ug/dL was considered normal. A UFC more than 3 times the upper reference level was likely to be CS. However, a UFC level between 1 and 3 times the upper reference level could be considered CS if other biochemical tests and the clinical picture clearly indicated CS. The hormonal follow-up pro- tocol after 24 months was identical to the initial biochemi- cal evaluation. However, in some cases with equivocal bio- chemical results or at the discretion of the physician, the hormonal follow-up was extended for an additional period. Simultaneous with retrieval of data from the electronic med- ical record system, the National Population Register was consulted to determine whether the person was still alive.
The study was approved by the Ethics Committee of the Karolinska Institutet, Stockholm, Sweden.
Statistical Analysis
Results for continuous variables are reported as the mean and standard deviation if normally distributed, oth- erwise as median and range. For categorical variables, frequencies are reported. Continuous data were com- pared using unpaired t tests for normally distributed data;
otherwise, the Mann-Whitney U test was used. The chi- square test was used for categorical data. Missing values were discounted when calculating proportions. A P value <. 05 was considered indicative of statistical significance.
RESULTS
Basic Characteristics of the Patients
The mean age of the 637 patients (403 females) diag- nosed with an AI was 62.7 ± 11.6 years (range, 21 to 89 years).
CT scanning was primarily used to discover AIs (n = 570, 92.2%). Thirty-five patients (5.5%) had been diagnosed by MRI, and 8 patients (1.3%) were diagnosed by ultrasonography (all subsequently confirmed by CT scan). The mean tumor size was 25.3 + 17.0 mm (range, 10 to 150 mm). Most AIs were unilateral (n = 563, 88.4%). Unilateral AI was more frequently found in the left adre- nal than in the right adrenal (348 [61.8%] cases versus 215 [38.2%] cases; P <. 001).
Hormonal Evaluation
The vast majority of the adrenal lesions were hormon- ally normal (91.4%). In 55 cases (8.6%), hormonal distur- bances were diagnosed during the study, most commonly SCS (Fig. 1). Hormonally active tumors were larger than nonfunctional tumors (39.3 ± 24.3 mm versus 23.9 ± 15.6 mm; P <. 001).
Two patients were diagnosed with congenital adre- nal hyperplasia (CAH) following analysis of urine ste- roid profile during the work-up for a case of suspected adrenocortical cancer (ACC, n=1) and during analysis of serum 17OHP levels in a female with bilateral AI and hir- sutism. These cases have been described previously (5,6). However, only 47 patients (7.4%) had 17OHP tested, and 26 (4.1%) had a urinary steroid profile measured; hence, in total, 68 patients (10.7%) had either or both measured. Thus, of those with an AI and steroid precursors measured, 2.9% were diagnosed with CAH. DST was performed in only 332 cases (52.1%). Inadequate suppression of cortisol after the DST was found in half of the cases (Table 1). There
| Table 1 Serum Cortisol Level (ug/dL) After 1-mg Dexamethasone Suppression Test (DST) | |
|---|---|
| Cortisol level after DST (ug/dL) | Number of patients (%) |
| <1.8 | 181 (54.5) |
| 1.8-2.2 | 33 (9.9) |
| 2.2-3.6 | 68 (20.5) |
| 3.6-5.0 | 18 (5.4) |
| >5.0 | 32 (9.6) |
were no differences between the normal and abnormal DST groups with respect to the size of the tumor and sex or age of the patient (data not shown). Among the patients with abnormal DST, 25 also had an ACTH level measured: 12 had suppressed ACTH, 12 had normal ACTH, and ACTH was elevated due to ectopic ACTH secretion of a small- cell lung cancer in 1 patient. Inadequate DST (30 of 31, 96.7%), low morning ACTH level (12 of 31, 38.7%), and elevated 24-hour UFC (5 of 31, 16.1%) were the most fre- quent criteria leading to a diagnosis of cortisol hypersecre- tion (26 SCS and 5 overt CS). Of the patients with a nor- mal DST result at baseline, all remained within the normal ranges during follow-up.
The majority of patients diagnosed with pheochro- mocytoma had elevated blood pressure (5 of 9, 55.6%) and were treated with 3.0 ± 1.4 blood pressure-lower- ing medications. None of these patients had complained of classic symptoms of pheochromocytoma prior to the finding of AI; however, when specifically questioned dur- ing the work-up for AI, most were found to be symptom- atic (Table 2). The reasons for performing CT scanning were: pain in the neck (n = 2), hematuria (n = 1), dyspnea (n = 1), urine retention (n = 1), acute diverticulitis (n = 1), ileus (n = 1), suspected aorta dissection (n = 1), and unclear reason (n = 1). The mean tumor size was 4.4 ± 2.2 cm. All patients diagnosed with pheochromocytoma underwent adrenalectomy, and the histologic examinations confirmed pheochromocytoma.
Metabolic Findings
Plasma glucose level and blood pressure were mea- sured in the vast majority of patients. Metabolic char- acteristics of the patients with SCS, overt CS, and
nonfunctioning AI are shown in Table 3. The prevalence of type 2 diabetes mellitus (T2DM) was similar in patients with SCS and those with nonfunctioning AI, whereas the prevalence of hypertension was higher in the SCS group. Especially in middle-aged patients (50 to 70 years old), hypertension was almost twice as common in the SCS group as in the adenoma group (P = . 044, data not shown). Unfortunately, there were no data on obesity and osteoporosis.
Bilateral Tumors
Bilateral adrenal tumors were discovered in 11% of patients, with higher rates in females than in males. Patients with bilateral tumors were of similar age as those with unilateral tumors (Table 4). Only the largest diameter of the adrenal lesion was registered, yet the bilateral lesions were larger than the unilateral lesions. The prevalence of T2DM in patients with bilateral tumors tended to be higher. Hormonally active tumors tended to be more frequent in the bilateral tumor group (14.3% versus 7.8%; P = . 072), the most prevalent hormonal abnormality in the bilateral tumor group being SCS.
Carcinomas
Four cases of ACC were detected; all were males. The age was 66.0 ± 5.3 years, compared to 62.9 + 11.4 years in patients with nonfunctional AI (P <. 001). The mean tumor size was 91 + 34 mm (range, 65 to 140 mm), which was greater than the mean nonfunctioning adenoma size (P = .001). In 2 cases, the tumors were hormonally active, with cortisol and androgen hypersecretion. Adrenalectomy was performed in all cases, and the diagnosis was confirmed histologically.
| Table 2 Data Regarding Patients Diagnosed with Pheochromocytomaª | ||||||
|---|---|---|---|---|---|---|
| No. | Symptoms | HT | Plasma metanephrine (nmol/L) | Plasma normetanephrine (nmol/L) | Urinary adrenaline (nmol/24 h) | Urinary noradrenaline (nmol/24 h) |
| 1 | Irritation | Yes | 5.2 | 5.5 | 493 | 335 |
| 2 | Anxiety, tremor, loss of weight | Yes | 3.4 | 5.1 | 234 | 674 |
| 3 | Headache | Yes | 0.4 | 1.0 | 127 | 610 |
| 4 | Irritation, palpitations | No | 0.4 | 1.6 | 83 | 1,130 |
| 5 | Palpitations | No | ND | ND | 40 | 897 |
| 6 | No symptoms | No | 0.4 | 7.0 | 32 | 717 |
| 7 | Dizziness | No | 0.6 | 0.7 | 87 | 783 |
| 8 | Palpitation attacks | Yes | 0.4 | 11 | 138 | 9,440 |
| 9 | No symptoms | Yes | 6.4 | 11 | 2,221 | 5,119 |
Abbreviations: HT = hypertension; ND = not determined; No. = patient number.
a Reference levels: plasma metanephrine, <0.3 nmol/L; plasma normetanephrine, <0.6 nmol/L; urinary adrenaline, <80 nmol/24 h; urinary noradrenaline, <400 nmol/24 h.
| NFA (n = 544) | SCS (n = 26) | CS (n = 5) | P value | |
|---|---|---|---|---|
| Age (years) | 62.92 ± 11.43 | 61.6 ± 11.05 | 53.80 ± 11.37 | .178 |
| T2DM, n (%) | 85 (15.3) | 6 (23.1) | 1 (20) | .581 |
| HT, n (%) | 264 (47.6) | 16 (61.5)ª | 4 (80) | .168 |
| DST ± SD | 2.16 ± 2.25 | 6.89 ± 4.97ª | 10.58 ± 10.65ª | .0001 |
| UFC ± SD | 26 ± 47.4 | 46 ± 50.2ª | 72.8 ±75.1ª | .012 |
Abbreviations: CS = Cushing syndrome; DST = cortisol value after dexamethasone suppression test (ug/dL); HT = hypertension; NFA = nonfunctioning cortical adenoma; SCS = subclinical Cushing syndrome; T2DM = type 2 diabetes mellitus; UFC = urinary free cortisol (ug/24 h).
a P <. 05 compared with the NFA group.
Fourteen patients (2.2%) were diagnosed with metas- tasis to the adrenal glands. The mean tumor size was 39.1 ± 23.0 mm, and the mean age of the patients was 68.6 ± 10.0 years. In 10 cases (71.4%), the tumor was unilateral. The primary origins of the malignancies were: lung (n = 5), colon (n = 2), breast (n =2), melanoma (n=1), B-cell lymphoma (n = 1), prostate (n = 1), hepatic (n = 1), and unknown (n = 1).
Follow-up
A total of 593 patients (93%) were followed for at least 24 months, according to the National Guidelines. In 42 cases (6.6%), information was missed due to death, patient refusal to be followed, or migration. However, survival data could be retrieved for all patients from the National Population Register. The mortality rate during the first year was 2.2%, and the rate was 12.4% over the entire period of up to 11 years from presentation of the AI to the date of data retrieval (February 2013). Of the 14 patients with an adrenal metastasis, only 1 was alive at 2 years from pre- sentation of the AI. Of the 4 patients diagnosed with ACC, 2 died within 27 months, and 1 patient died after 7 years of tumor burden. One patient was still alive 7 years after the diagnosis. The causes of death in all other cases were considered unrelated to adrenal disease. The mortality rate due to cardiovascular diseases as a result of long-term mild cortisol hypersecretion was not investigated.
In 550 patients (86.3%), the size of tumor had not changed by the time of the last follow-up. In 34 cases (5.3%), the tumor had grown more than 0.5 cm per year, and all of these patients underwent adrenalectomy. The final diagnosis after adrenalectomy was: nonfunctional adenoma (n = 27), SCS (n = 4), CS (n = 1), adrenal metas- tasis (n = 1), and myelolipoma (n = 1).
In 285 of 332 cases (85.8%) in which a DST was per- formed at baseline, the size of the tumor remained the same
during the follow-up period. In 18 cases (5.4%), the tumor had grown more than 0.5 cm, and in 1 case (0.3%), the tumor had decreased in size. In 27 cases (8.1%), there were no data regarding tumor size.
Almost all (99.6%) of the nonfunctioning AIs remained hormonally inactive during the follow-up period. However, 2 patients with initially normal hor- monal values developed SCS during the follow-up. The first patient was a 57-year-old obese woman suffering from posttraumatic epilepsy, hypertension, obstructive sleep apnea, and T2DM (diagnosed 6 months before the CT scan). The imagining was due to suspected hepatitis, and a 40-mm AI was revealed. The initial morning serum and urine cortisol levels were normal; however, no DST was performed. A DST performed at the 2-year follow- up revealed insufficient suppression (13.8 ug/dL), but her UFC remained normal. Adrenalectomy was subsequently done. The second patient was a 55-year-old woman who presented with soreness in the abdomen without any clinical signs of hypercortisolism. The CT scan revealed a 17-mm AI. Morning serum and urinary cortisol levels were normal initially, but no DST was performed. After 1 year, her UFC remained normal; however, an abnormal DST (6.5 µg/dL) result was obtained, and the patient was diagnosed with SCS. Both patients were alive at end of the study retrieval.
Adrenalectomy
Adrenalectomy was performed in 46 patients (7.2%). The indications for surgery were: radiologic suspicion of ACC (n = 17, of which only 4 were ACC), primary hyper- aldosteronism (n = 3), pheochromocytoma (n = 9), SCS with severe comorbidities (n = 12), CS (n = 4), and adrenal metastasis (n = 1). The mean size of surgically removed AIs was 45.4 ± 24.3 mm. Most of the SCS patients (9 of 12, 75%) required postoperative cortisol replacement.
| Bilateral (n = 70, 11.0%) | Unilateral (n = 563, 88.2%) | P value | |
|---|---|---|---|
| Age (years) | 64.7 ±9.7 | 62.5 ±11.8 | .133 |
| Female | 74.3% | 61.6% | .048 |
| Size (mm)ª | 29.9 ± 18.0 | 24.8 ± 16.9 | .005 |
| Hypertension, n (%) | 37 (52.9) | 270 (48.0) | .228 |
| T2DM, n (%) | 15 (21.4) | 81 (14.4) | .066 |
| Type of tumor | .015 | ||
| NFA, n (%) | 53 (75.7) | 489 (86.9) | .018 |
| SCS, n (%) | 7 (10.0) | 19 (3.4) | .018 |
| Cushing syndrome, n (%) | 2 (2.9) | 3 (0.5) | .097 |
| Primary hyperaldosteronism, n (%) | 0 (0) | 9 (1.4) | .607 |
| Pheochromocytoma, n (%) | 0 (0) | 9 (1.4) | .607 |
| ACC, n (%) | 1 (1.4) | 3 (0.5) | >.999 |
| CAH, n (%) | 0 (0) | 2 (0.4) | >.999 |
| Otherb, n (%) | 3 (4.3) | 20 (3.6) | .733 |
Abbreviations: ACC = adrenal cortical carcinoma; CAH = congenital adrenal hyperplasia; NFA = nonfunctioning cortical adenoma; SCS = subclinical Cushing syndrome; T2DM = type 2 diabetes mellitus. a Size of the largest tumor.
b Myelolipoma, cyst, or adrenal bleeding.
DISCUSSION
In this large cohort of 637 AI patients, we demonstrated that the vast majority of AIs were nonfunctional adenomas. The most common hormonal abnormalities were subtle signs of suppression of the HPA axis driven by low active autonomous cortisol secretion (i.e., SCS, accounting for one in every 25 AIs). Of those patients screened for 17OHP or urine steroid profile (of which there were few), almost 3% were diagnosed with CAH. ACC was rare, but patients with adrenal metastases were more common and had an extremely poor prognosis, as almost all had died within 1 year. The initial evaluation diagnosed in principal all func- tional adenomas.
More than a half of our patients were women (63.3%), the mean age was 63 years, and there was an increasing frequency of AIs with age, findings that were similar to those of previous studies (7-9). In contrast, no correlation between age and tumor size was found. Hormonally active lesions were larger than nonactive lesions, which has also been reported by others (4,10). Eighty-five percent of the AIs were nonfunctioning adenomas, and the most frequent hormonal disorder was subclinical or overt autonomous
glucocorticoid hypersecretion, similar to other reports (4,10,11). However, there is no consensus regarding how to diagnose and treat SCS (3,7,12). We diagnosed SCS based on the presence of at least 2 abnormal biochemical tests involving the HPA axis. Patients with HPA axis aberrations had larger tumors and more often bilateral tumors. Olsen et al (13) observed an abnormal DST in 70% of patients with bilateral AI, and 42% were diagnosed with SCS. Similarly, in our study, the frequency of SCS was significantly higher and T2DM tended to be more prevalent in the group with bilateral tumors; however, only 10% were considered to have SCS. The reason for this prevalence difference is unclear but may be related to different diagnostic criteria. Moreover, many patients did not have a DST, which may have underestimated the frequency. On the other hand, a DST with a low plasma cortisol cutoff <1.8 ug/dL is asso- ciated with a high rate of false-positive cases and a low positive predictive value. In our study, 45.5% of patients had a positive DST, but only 5% had hypercortisolism dis- orders. It is known that factors such as advanced age, obe- sity, rapid elimination of dexamethasone, stress disorders such as depression, and alcohol overconsumption can cause false-positive test results. However, our SCS group had a
higher prevalence of hypertension compared to patients with nonfunctioning AIs. Half of the SCS group underwent an adrenalectomy, and of those, 75% needed postoperative glucocorticoid replacement due to a suppressed contralat- eral adrenal gland, indicating a correct diagnosis.
Mantero et al (10) found that clinically indolent pheo- chromocytoma is the second most common form of hyper- functioning AI, with a prevalence of 4.2%, half of which are normotensive. In our study, the prevalence was only 1.4%, but the frequency of normotensive cases was the same. The difference in occurrence reflects either referral bias or a lower prevalence of pheochromocytoma in our catchment area. We excluded all pheochromocytomas that had presented to other centers as AIs and thereafter were referred to our center in order to establish the pheochromo- cytoma diagnosis. As pheochromocytoma is a potentially lethal condition, it is essential to identify these cases. All of our cases underwent successful adrenalectomy.
Mantero et al (10) also reported enhanced levels of 17OHP after ACTH stimulation testing. This was found in half of the patients with cortical tumors screened with- out regard to differences between malignant or benign lesions. In our study, 17OHP was measured randomly in a few patients, and the urinary steroid profile was deter- mined in those with suspicion of ACC. This revealed 2 cases of CAH out of 68 cases in which steroid metabolites were measured, resulting in a CAH prevalence of 2.9%. Moreover, being a carrier or having CAH may lead to an increased susceptibility to adrenal tumors (5,6,14). This suggest more frequent measurement of 17OHP (especially stimulated 17OHP) is necessary, as the basal 17OHP level may be normal despite increased stimulated 17OHP levels (6). In a study involving 50 patients with AI, Baumgartner- Parzer et al (15) found 1 patient with CAH and 8 carri- ers. Hence, 18% had at least one mutation in the CYP21A2 gene, the gene associated with most forms of CAH. In contrast, in 202 patients with AI described by Barzon et al (16), only 1 patient (0.5%) was diagnosed with CAH. Thus, more studies are needed to clarify whether screening for CAH or carrier status should be done routinely.
The frequency of ACC among AI varies from 1.2 to 12%, depending on the study (7). We found a frequency of only 0.6%. This may be explained by the fact that the more specialized the center, the higher the frequency of ACC. However, our center is a major tertiary center, and so it should be subject to the same selection bias as other large centers. We had many more ACC cases referred, but we only included those involving presentation as an AI, which may differ from previous studies. None of our ACCs was smaller than 65 mm; thus, the recommendations for surgi- cal treatment of tumors larger than 4 cm in diameter (17) seems reasonable.
According to the literature, the prevalence of adrenal metastasis varies from 0.3 to 11.1% (4,10,18). In a study
of unknown primary cancers, involvement of the adrenal gland at presentation was demonstrated in 5.8% of cases, and a single metastasis limited only to the adrenals was observed in 0.2% (19). However, none fulfilled the criteria for a true AI. In our study, the frequency of adrenal metas- tasis was 2.2%, regardless of the presence of metastasis to other organs. Yet, in our study, we excluded patients with indications of staging or control of a known malignancy, and all patients were evaluated due to other reasons. Lee et al (19) demonstrated an extremely poor prognosis similar to ours in cancer patients with adrenal involvement.
If patients with ACC and adrenal metastasis were excluded, survival appeared normal. Considering the mean age of 63 years at diagnosis of a nonmalignant AI, 2.2% had died after 1 year, and 12.4% had died by the end of the study, some with up to 11 years of follow-up. This contradicts the results of Barry et al (20), who reported a mortality rate of 36% (81 of 224 AI patients) over a mean follow-up of 7 years. Moreover, only 2 of our patients were diagnosed with a hormonal overproduction at the follow-up visit, both of whom had SCS, but unfortunately, neither of them had had a DST performed at initial work- up. Thus, a follow-up of 2 years seems unwarranted if the initial evaluation does not reveal any hormonal deviations or radiologically suspicious signs. This is similar to the results published by Muth et al (21). We suggest no further follow-up for adrenal tumors ≤4 cm with benign radiology and a normal hormonal profile at the initial work-up.
CONCLUSION
In conclusion, even though the vast majority of AIs in this study were benign nonfunctioning adenomas, 11% of cases involved hormonally active or malignant lesions, thus underscoring the importance of adequate biochemi- cal and radiologic examinations of the patient at the ini- tial work-up. However, only 2 patients (0.3%) developed hormonal overproduction during follow-up, and these 2 cases of SCS would probably have been found if an initial DST had been performed. No additional malignancy was found during follow-up. The prognosis of those patients diagnosed with adrenal metastasis was extremely poor, but otherwise, the prognosis was normal in patients with AI. Hence, follow-up of AI with an initial nonfunctional pro- file and benign radiologic appearance seems unnecessary.
ACKNOWLEDGMENT
This study was supported by the Magn. Bergvalls Foundation, Karolinska Institutet, and Stockholm County Council.
DISCLOSURE
The authors have no multiplicity of interest to disclose.
REFERENCES
1. Androulakis II, Kaltsas G, Piaditis G, Grossman AB. The clinical significance of adrenal incidentalomas. Eur J Clin Invest. 2011;41:552-560.
2. Mazzuco TL, Bourdeau I, Lacroix A. Adrenal inciden- talomas and subclinical Cushing’s syndrome: diagnosis and treatment. Curr Opin Endocrinol Diabetes Obes. 2009; 16:203-210.
3. Terzolo M, Stigliano A, Chiodini I, et al. AME position statement on adrenal incidentaloma. Eur J Endocrinol. 2011;164:851-870.
4. Bülow B, Ahren B. Adrenal incidentaloma-experience of a standardized diagnostic programme in the Swedish pro- spective study. J Intern Med. 2002;252:239-246.
5. Falhammar H, Thorén M. An 88-year-old woman diag- nosed with adrenal tumor and congenital adrenal hyperpla- sia: connection or coincidence? J Endocrinol Invest. 2005; 28:449-453.
6. Falhammar H. Non-functioning adrenal incidentalomas caused by 21-hydroxylase deficiency or carrier status? Endocrine. 2014;47:308-314.
7. Arnaldi G, Boscaro M. Adrenal incidentaloma. Best Pract Res Clin Endocrinol Metab. 2012;26:405-419.
8. Bülow B, Jansson S, Juhlin C, et al. Adrenal inciden- taloma - follow-up results from a Swedish prospective study. Eur J Endocrinol. 2006;154:419-423.
9. Hammarstedt L, Muth A, Wängberg B, et al. Adrenal lesion frequency: a prospective, cross-sectional CT study in a defined region, including systematic re-evaluation. Acta Radiol. 2010;51:1149-1156.
10. Mantero F, Terzolo M, Arnaldi G, et al. A survey on adre- nal incidentaloma in Italy. J Clin Endocrinol Metab. 2000; 85:637-644.
11. Comlekci A, Yener S, Ertilav S, et al. Adrenal inciden- taloma, clinical, metabolic, follow-up aspects: single centre experience. Endocrine. 2010;37:40-46.
12. Zeiger MA, Thompson GB, Duh QY, et al. The American Association of Clinical Endocrinologists and American
Association of Endocrine Surgeons medical guidelines for the management of adrenal incidentalomas. Endocr Pract. 2009;15:1-20.
13. Olsen H, Nordenström E, Bergenfelz A, Nyman U, Valdemarsson S, Palmqvist E. Subclinical hypercor- tisolism and CT appearance in adrenal incidentalomas: a multicenter study from Southern Sweden. Endocrine. 2012;42:164-173.
14. Jaresch S, Kornely E, Kley HK, Schlaghecke R. Adrenal incidentaloma and patients with homozygous or hetero- zygous congenital adrenal hyperplasia. J Clin Endocrinol Metab. 1992;74:685-689.
15. Baumgartner-Parzer SM, Pauschenwein S, Waldhäusl W, Pölzler K, Nowotny P, Vierhapper H. Increased prevalence of heterozygous 21-OH germline mutations in patients with adrenal incidentalomas. Clin Endocrinol. 2002;56:811-816.
16. Barzon L, Scaroni C, Sonino N, Fallo F, Paoletta A, Boscaro M. Risk factors and long-term follow-up of adre- nal incidentalomas. J Clin Endocrinol Metab. 1999;84: 520-526.
17. Arnaldi G, Masini AM, Giacchetti G, Taccaliti A, Faloia E, Mantero F. Adrenal incidentaloma. Braz J Med Biol Res. 2000;33:1177-1189.
18. Bencsik Z, Szabolcs I, Góth M, et al. Incidentally detected adrenal tumours (incidentalomas): histological heterogene- ity and differentiated therapeutic approach. J Intern Med. 1995;237:585-589.
19. Lee JE, Evans DB, Hickey RC, et al. Unknown primary cancer presenting as an adrenal mass: frequency and impli- cations for diagnostic evaluation of adrenal incidentalomas. Surgery. 1998;124:1115-1122.
20. Barry MK, van Heerden JA, Farley DR, Grant CS, Thompson GB, Ilstrup DM. Can adrenal incidentalomas be safely observed? World J Surg. 1998;22:599-604.
21. Muth A, Hammarstedt L, Hellström M, Sigurjónsdóttir HÁ, Almqvist E, Wängberg B. Cohort study of patients with adrenal lesions discovered incidentally. Br J Surg. 2011;98:1383-1391.