Adrenal Incidentaloma
Brandi Jones Salomone, MD
Resident Grand Rounds
May 21, 2002
Adrenal masses discovered by imaging studies performed for unrelated reasons have become a common clinical problem. Before 1974, the literature reports 178 cases of nonfunctioning adrenal cortical tumors (1). Since that time, the introduction of computed tomography (CT) and magnetic resonance imaging (MRI), have resulted in the identification of clinically silent adrenal masses as small as 5 mm in diameter in 1-10% of patients studied for reasons other than suspected adrenal disease (2). This evidence is in accordance with data from autopsy series documenting microscopic or macroscopic adrenal nodules in 2-9% of patients who had never had clinical signs of adrenal dysfunction during life (3). When detected, clinically in-apparent adrenal masses raise challenging questions for physicians and their patients regarding evaluation and treatment. The purpose of this paper and presentation is to help the clinician make a decision on how to properly evaluate and treat an adrenal incidentaloma (AI) based on the current available evidence. In addition, the NIH Consensus Statement Guidelines (February 2002) for the evaluation and treatment of AI will be reviewed.
The adrenal glands sit atop each kidney in the retroperitoneal space and are comprised of the central medulla and exterior cortex. The adrenal cortex functions as the body’s main producer of cholesterol-based steroids which include androgens, glucocorticoids and mineralocorticoids while the adrenal medulla produces stores and secretes catecholamines.
Figure 1 Biosynthetic pathways for adrenal steroid
production.
The first question a clinician must ask when evaluating an
adrenal incidentaloma is, “Is it hormonally active?” Any lesion of the adrenal gland has the
potential to cause a hyper-secretory state, however most adrenal incidentalomas
(AIs) do not overproduce hormones. The
overproduction syndromes associated with an AI are pheochromocytoma, primary
aldosteronism, virilization/feminization and Cushing’s syndrome. More
recent reports suggest up to 20% of patients with an AI have some form of sub-clinical
hormonal dysfunction (4). In addition to a hyper-secretory tumor the
differential for an AI includes benign adenoma, adrenocortical carcinoma (ACC),
metastasis (usually from the breast, lung, kidney or colon), myelolipoma and pheochromocytoma.
Adrenal
cortex
Adenoma
Nodular
hyperplasia
Carcinoma
Adrenal
Medulla
Pheochromocytoma
Ganglioneuroma
Ganglioneuroblastoma
Other adrenal
masses
Myelolipoma
Table
1 Incidentally discovered adrenal masses.
Neurofibroma
Hamartoma
Teratoma
Xanthomatosis
Amyloides
Cyst
Hematoma
Granulomatosis
Metastases
Breast
carcinoma
Lung
cancer
Lymphoma
Leukemia
Renal
carcinoma
The second question a clinician must answer is “Is the AI benign or malignant”. Fortunately ACC accounts for only 2% of AIs < 4 cm, 6% of AIs between 4.1-6.0 cm, and 25% of AI > 6 cm (4). The answers to these questions provide the clinician a framework to guide further treatment of an adrenal incidentaloma.
In a large retrospective cohort study published in Surgery in 1991, Herrera, et al at the Mayo Clinic developed guidelines for investigation and management of adrenal incidentalomas. They sought to answer: (1) do AIs need to be evaluated routinely by endocrine functional studies, (2) is there a size below which nonfunctioning masses can be observed safely, and (3) how should patients with incidentalomas under observation be evaluated subsequently? They defined AI as “a tumor serendipitously discovered by radiologic examination, in the absence of symptoms or clinical findings suggestive of adrenal disease”(5). His group collected data from 61,054 CT scans done at their institution between 1985 and 1989. Of those, 2066 (3.4%) were found to have an adrenal mass. They excluded patients for the following reasons: (1) previous or concurrent malignancy, (2) adrenal tumors localized after biochemical documentation of disease, (3) adrenal nodules < 1 cm, (4) a patient with localized tumor related symptoms or general constitutional symptoms strongly suggestive of malignant disease, and/or (5) evidence of local invasion. Patients with previously diagnosed cancer were excluded because approximately two-thirds of clinically in-apparent adrenal masses are metastases in that population (4). From these criteria, 259 patients were enrolled in the study as true AI. An additional 89 patients referred to Mayo Clinic with CT scan in hand were included in the study group for a total of 342 patients. The following pie graph shows the breakdown of the 2066 adrenal masses. AI made up 16.5%.
Figure 2
Distribution of 2066 adrenal abnormalities based on clinical
factors. Herrera.
Of the 342 cases, 136 were men and 206 were women with a mean age of 62. Reasons for CT scan included nonspecific abdominal pain, known abdominal or thoracic disease, renal problems, musculoskeletal back pain as well as other miscellaneous symptoms. Of the 342, one hundred seventy, roughly half of the patients, underwent studies for hyper-function including urinary metanephrines or VMA for catecholamine excess, urinary excretion of 17-hydroxycorticosteroids, serum concentration of cortisol or 1 mg dexamethasone suppression test for glucocorticoid excess and serum potassium for mineralocorticoid excess. No further hormonal evaluation was performed for androgen hyper-secretion since no patients exhibited any feminizing or virilizing characteristics. According to this study, of the 170 cases chosen to undergo hormonal evaluation, five were found to have pheochromocytomas, only one of which had hypertension, and two had cortisol producing adenomas. Furthermore, via follow-up by personal interview one to three years later, 287 of the 342 showed no clinical evidence of hormonal excess, however, none of these patients underwent hormonal investigation to confirm this. To summarize, only 4.1% of those tested had evidence of hormonal over-production. With the entire study population included in the calculation the percentage with hyper-secretion decreases to 2.8%. From this data Herrera et al conclude that biochemical screening for a hyper-secretory state should be performed on patients with AIs > 1 cm in size, however, no where in the article do they clarify their own criteria for testing patients for hyper-secretion. They also conclude that any patient with hormone over-production should undergo adrenalectomy.
In this study, 52 of the 342 patients underwent adrenalectomy. Pathologic analysis confirmed five malignancies. Four of these were adrenocortical carcinoma and one was metastasis. Table 2 summarizes the hyper-secreting lesions and the malignant tumors.
Table 2
Characteristics of patients with functioning and malignant tumors. Herrera.
None of the malignancies were less than 5 cm in diameter confirming previous data from the 1980’s that ACC prevalence is much higher in larger tumors. In this study malignancy makes up only 1.2% of the 342 cases. Despite this low prevalence, the authors conclude that surgical excision of the affected adrenal gland should be offered to anyone with an AI > 4cm, however, no where in the study do they discuss why these fifty two patients underwent adrenalectomy. Ninety percent of the 287 that followed up were seen at the one year mark. One hundred and fifty-nine (55%) of these patients had a repeat CT scan that did not show any development of tumor extension. Despite surgery, four of the five patients with malignancy were dead at follow-up questioning whether surgery prolongs survival. In a previous clinicopathologic study of 42 patients with histologically proven ACC, Didolkar et al determined the overall median and five year survival rates to be 14 months and 24% (6). Lastly, Herrera et al conclude that comparison CT scans should be done on patients with AIs <4cm choosing observation at three month intervals. However, 90% of the 287 that followed up in this study were re-evaluated at one year questioning this recommendation.
In 1997 Kasperlik-Zaluska et al published a study in Clinical Endocrinology summarizing their experience in the diagnosis, pathological findings and treatment of a large group of patients with AI (7). Over a ten year period 208 patients were referred to their institution for further investigation of an adrenal mass discovered incidentally on ultrasound performed for one of the following reasons: abdominal pain, lumbar pain, nephrolithiasis, urinary tract infection, hematuria or erythrocytosis. As part of the inclusion criteria each patient underwent a CT scan that confirmed the AI. Of the 208 patients one hundred forty eight of the cases were women and 60 were men. The researchers attempted to exclude patients with malignancy by performing oncologic screening tests including chest roentgenogram, a digital rectal exam, a clinical breast exam, a pap smear, a CDP and an ESR. None of these patients had positive screening tests.
In patients presenting with hypertension or arrhythmia, twenty four hour urinary excretion of 17-hydroxycorticosteroids (17-OHCS) and urinary catecholamines were performed. One hundred seven patients had am cortisol and ACTH levels drawn , 89 of which also had the same tests at night. The researchers failed to clarify why these patients were selected for testing. Forty eight patients with some clinical evidence of Cushing’s syndrome, obesity, HTN, DM or those with a high normal 17-OHCS underwent low dose dexamethasone suppression testing. 17-OHCS levels were elevated in two patients (4.17% of those tested; 0.92% of total patients in study) diagnosed with Cushing’s disease and eventually adrenal adenomas. Three patients diagnosed with ACC also had elevated 17-OHCS urinary excretion (6.3% of those tested; 1.4% of total patients in study). Of the forty eight patients tested for 17-OHCS urinary excretion, five had elevated levels (10.4%) which accounts for only 2.4% of the entire study population. Nine patients had increased urinary catecholamines but it is not stated how many people were tested for this possible abnormality. All the patients with evidence of hormonal excess underwent adrenalectomy. In addition anyone presenting with a tumor >4 cm was advised to have surgery and 59 gave informed consent. This research team designated their cutoff size based on surgical data, such as Paul Copeland’s review in 1983 of over 50 articles on AI, in which he found that 105 of 114 documented ACCs were over 6 cm in diameter (2). Thus, 85 patients of the original 208 underwent adrenalectomy because of an AI > 4 cm and/or existence of a hyper-secretory state. The breakdown of the pathologic diagnosis postoperatively was 21 adenomas, 17 ACC, 13 pheochromocytomas, 12 metastasis, 10 myelolipomas and 12 others. Figure 3 shows the distribution of the various lesions according to size.
The malignant tumors ranged from 3.2 to 20 cm but only one was less than 5 cm, supporting Copeland’s observation that AIs less than 3-4 cm are rarely malignant. However, the authors also point out that the benign tumors ranged from 1.5 cm to 21 cm leaving a large area for overlap. Even still, Kasperlik-Zaluska et al conclude that size proved poorly specific but highly sensitive when used as a predictor of malignancy. In this study ACC is highly prevalent, accounting for 21% of the patients undergoing adrenalectomy. A diagnosis of ACC holds a poor prognosis. According to the NIH consensus statement released in February 2002, “the overwhelming majority of studies report less than 50 percent 5-year survival rates for adrenal cortical carcinoma, and several report less than 50 percent 2-year overall survival” (4). Also, there is no evidence that early detection of ACC decreases mortality. Of the 208 initial cases, the prevalence of ACC decreases to 8.6%. The patients going for surgery were carefully selected based on size criteria and hormonal hyper-secretion. The authors summarize that unnecessary surgery can be avoided in patients with AI if the criteria for adrenalectomy are tumor size >4 cm and/or the presence of a hyper-secretory state. They recommend endocrinological screening in those with clinical symptoms worrisome for hyper-function and/or those electing surgery based solely on tumor size because of the peri-operative mortality associated with an untreated pheochromocytoma. A source of bias in the study is that all the patients were referred secondary to an AI found on ultrasound, capable of finding only larger lesions which are more at risk of being cancer. Another weakness is a lack of consistent long term follow-up despite recommending that those choosing observation only should have repeat ultrasounds at 1, 3 and 6 month intervals. This study lacks any clear evidence for these follow-up intervals.
Luton et al published a retrospective cohort study in the European Journal of Endocrinology in
2000 looking at the outcomes of 88 patients with an AI who underwent
adrenalectomy and an additional four that had histologically proven diagnoses (8). Patients had abdominal CT scans performed
for many reasons including: medical check-up for a non-endocrine disorder
(31%), pain syndrome (16%), gastrointestinal symptoms (14%), urinary symptoms
(11%), hepatic symptoms (7%), and others (21%).
Of the 208 patients referred to the Hospital Cochin for an adrenal mass
from 1978 to 1996, one hundred eighty underwent a complete physical exam, an
abdominal CT scan and multiple endocrinological studies. The article mentions no exclusion criteria
for the twenty eight patients not enrolled in the study. The
following table illustrates the endocrine investigations.
Table 3 Endocrinological studies.
