5.3 Adrenal incidentaloma

Adrenal incidentaloma is an adrenal mass that is discovered serendipitously with a radiological examination performed for indications unrelated to adrenal disease (1). The incidental discovery of an adrenal mass has become an increasingly common problem, because of the widespread use of ultrasonography, CT, and MRI in clinical practice (2, 3). These techniques have greatly improved their power of resolution over recent years, thereby increasing the possibility of detection of tiny adrenal lumps.

Several factors hinder a clear characterization of the phenomenon ‘adrenal incidentaloma’, which may be considered as a byproduct of technology applied to medical practice. Adrenal incidentaloma is not a single pathological entity and the likelihood of any specific diagnosis depends both on the circumstances of discovery and the applied definition of incidentaloma. Unfortunately, published reports are inconsistent in applying inclusion and exclusion criteria for these various factors, making the results difficult to interpret. A further issue is the lack of specific clinical features of the patients carrying an adrenal incidentaloma.

In autopsy series, the mean prevalence of clinically inapparent adrenal masses is about 2.0%, ranging from 1.0 to 8.7% (4). This variability reflects different definitions and also the difficulty in distinguishing larger nodules within adrenal hyperplasia from distinct adrenocortical adenomas. The mean prevalence of adrenal incidentalomas in CT scan series published from 1982 to 1994 was 0.64%, ranging from 0.35 to 1.9% (4). More recently, we have found a frequency of benign adrenal masses of 4.2% in middle-aged subjects who were enrolled in a screening programme of lung cancer (5). The frequency of incidental adrenal masses is up to 4.4% in patients with a clinical history of cancer and 50–75% of adrenal nodules diagnosed in such patients are metastases, since the adrenal gland is frequently involved by metastatic spread. Many malignancies can metastasize to the adrenals, most frequently lung cancer, breast cancer, kidney cancer, melanoma, and lymphoma (4).

Adrenal incidentalomas show different distribution in the population dependent on the patient’s age and sex. In clinical reports, adrenal incidentalomas show a peak incidence between 50 and 60 years of age (4). This pattern could merely reflect a higher number of diagnostic procedures in these age decades or be the consequence of the ageing process of the adrenal glands, which may lead to increased formation of cortical nodules secondary to vascular changes (4). The frequency of adrenal incidentalomas is very low in childhood and adolescence (0.3–0.4% of all neoplasms in children). Unfortunately, the frequency of adrenocortical carcinoma within adrenal neoplasms in children is very high, about 80%. Adrenal cancer represents 1.3% of all malignancies in patients less than 20 years and frequency is higher in children under 6 years (4).

The sex distribution is characterized by a male to female ratio of 1:3 to 1:5. A higher prevalence of adrenal incidentalomas in women is likely to be partly explained by a referral bias (i.e. more imaging studies are done in women due to higher prevalence of biliary disease) as nonfunctioning adrenal adenomas occurred with comparable frequency in men and women in autopsy series (4).

Adrenal incidentalomas are more frequent in the Caucasian population and the right adrenal gland is affected in 50–60% of cases, the left one in 30–40%, while bilateral lesions are found in 10–15% (4). This right-side predominance reflects the fact that in most series adrenal masses were discovered by ultrasonography, which is less accurate in detecting masses on the left side (5). No side-related difference was reported in CT scan and autopsy series (4, 6).

Aetiology of adrenal incidentalomas includes either benign or malignant lesions. However, an adrenal incidentaloma is generally benign, being an adrenal adenoma in approximately 70% of cases (cortisol-secreting in 1–29%, aldosterone-secreting in 1.6–2.3%) (7). The frequency of phaeochromocytoma is estimated at 1.5–23% (7), that of adrenal cancer varies from 1.2% to 11% (7). The risk of an adrenal incidentaloma being an adrenal cancer is linearly related to the mass size, but this correlation is not apparent for metastases of extra-adrenal cancers (7). Other causes of adrenal incidentaloma are adrenal cysts, ganglioneuromas, myelolipomas, haematomas, and metastases of other malignancies. Moreover, adrenal lesions are found in inherited endocrine cancer syndromes (McCune–Albright syndrome, multiple endocrine neoplasia) and in insufficiently controlled congenital adrenal hyperplasia.

In a multi-institutional, retrospective survey performed in Italy including 1004 patients, of whom 380 underwent surgery, the most frequent pathological diagnoses were adrenocortical adenoma (52%), adrenocortical carcinoma (12%), phaeochromocytoma (11%), and myelolipoma (8%) (6). Obviously, the frequency of adrenocortical carcinoma and phaeochromocytoma was likely to be overestimated in this surgical cohort. Between 1991 and 2005, we collected a series of 181 patients at our institution and found adrenal adenoma to be the by far most frequent tumour type (Fig. 5.3.1). Establishing the precise aetiology of adrenal incidentaloma is difficult because surgical series have usually a selection bias towards masses that have a higher probability of being malignant or functioning, while series collected in medical departments have the limitation that most diagnoses are ascertained only by imaging and clinical criteria.

Adrenal incidentaloma is a growing public health challenge since the serendipitous detection of an adrenal mass increases with age and is expected to rise in populations that are getting older and have widespread access to ever improving radiological techniques (3). An impressive variety of tumoural and nontumoural lesions arising from the adrenal glands or extra-adrenal tissues may present as an adrenal mass detected serendipitously. Before embarking on a cumbersome diagnostic process, it is important to determine the most important questions the diagnostic work-up has to answer (Box 5.3.1). Following these concepts, it may be recommended to identify either primary adrenocortical carcinoma (ACC) or secondary adrenal malignancy (metastasis), and to rule out phaeochromocytoma (Fig. 5.3.2).

There is no doubt that ACC may significantly affect patients’ health and there is sufficient evidence to recommend surgery whenever possible. MacFarlane has reported that patients with untreated ACC have a median survival of 3 months (8), while complete surgical resection continues to be the treatment of choice for ACC and a margin-free resection is a strong predictor of long-term survival (7). The suspicion of ACC is raised by radiological criteria and finally verified by histopathology; fine-needle aspiration biopsy (FNAB) is currently not indicated for the diagnosis of primary ACC because of poor differentiation from adenoma and safety issues (9). The risk of ACC is related to the mass size, even if the correlation is far from perfect (Fig. 5.3.3). In the -multicentre Italian experience, a cut-off at 4 cm had the highest sensitivity to differentiate ACC from benign lesions. The positive predictive value, however, was low because benign lesions greatly exceeded the incidence of ACC at any tumour size (6).

Despite that, only a limited number of patients with ACC have been included in imaging studies, current criteria suggestive of a benign adenoma include attenuation values less than 10 Hounsfield units (HU) on unenhanced CT scans and less than 30 HU on enhanced scans. Tumours with more than 10 HU include adrenal adenomas with a low lipid content, phaeochromocytoma, metastasis, and ACC (see Chapter 5.1).

Recently, the analysis of the SEER database, a comprehensive national cancer registry compiled by the NCI, confirmed that an increased tumour size correlates with a higher likelihood of malignancy (10). The subset of data on 192 ACCs presenting with localized disease showed that a tumour size of 4 cm had a sensitivity of 96% and specificity of 52% for ACC (10), a figure very close to that observed in the Italian survey on adrenal incidentaloma (6). However, since the prevalence of malignant neoplasms is low among adrenal incidentalomas, the post-test probability of malignancy associated with any tumour size remains low (10).

It is also important to rule out phaeochromocytoma because it can lead to significant morbidity and mortality, particularly if it remains undiagnosed. An increasingly higher number of patients harbouring a phaeochromocytoma are found to be normotensive or have stable, -low-grade hypertension. In a large, multi-institutional series of adrenal incidentalomas collected in Italy and Sweden, approximately 50% of the patients bearing incidental phaeochromocytoma were normotensive or had stable, low-grade hypertension, which was indistinguishable from essential hypertension (6, 11). On the contrary, approximately 10% of the benign sporadic adrenal phaeochromocytomas diagnosed at the Mayo Clinic from 1978 to 1995 presented as incidentalomas. About 90% of these cases were -hypertensive and had diagnostic values of urinary catecholamines or metanephrines (12). Furthermore, in a multi-institutional survey performed in France, the frequency of incidentally detected phaeochromocytomas was 15.9% in the whole series, while the proportion of clinically silent phaeochromocytomas was as high as 25% among patients operated in more recent times (13). These findings confirm that phaeochromocytomas may present with mild symptoms, if any (4). In fact, several series reported that in a relevant number of cases phaeochromocytoma may only be discovered by autopsy (4).

In general, incidental phaeochromocytomas are large masses, greater than 4 cm (14). Large phaeochromocytomas are able to extensively metabolize catecholamines prior to secretion; they may therefore exhibit fewer clinical symptoms than small tumours (15). Unenhanced CT is accurate in the detection of adrenal phaeochromocytoma with a sensitivity ranging from 93 to 100% (16). Intravenous contrast enhancement is not generally essential even if most tumours enhance markedly after intravenous contrast medium (14). MRI may have higher specificity than CT in diagnosing adrenal phaeochromocytoma and is also more accurate in detecting the infrequent extra-adrenal phaeochromocytoma (17). T2-weighted MRI may be particularly helpful, as this tumour usually shows a high signal intensity (higher than ACC or metastasis) (4). Adrenal scintigraphy with [¹³¹I]metaiodobenzylguanidine (MIBG) provides anatomical localization and functional characterization of the tumour. This technique has lower sensitivity (78%) compared with CT or MRI but superior specificity (100%). MIBG scintigraphy may be useful in patients with equivocal imaging and biochemical data, or when a malignant or multifocal phaeochromocytoma is suspected (15).

Since nearly one-third of all phaeochromocytomas show a nonspecific appearance upon imaging, it is mandatory to perform an appropriate biochemical screening in every patient with an adrenal mass. Screening is of utmost importance whenever FNAB or surgical removal of the mass is scheduled (6). Prompt surgical resection remains the standard curative modality after specific preparation of the patient because up to 80% of patients with unsuspected phaeochromocytoma who underwent surgery or -anaesthesia have died (8).

The adrenal glands are a common site of metastasis, with a reported rate in patients with an extra-adrenal malignancy ranging from 32 to 73% in different series (8). The morphological CT imaging features of metastases are nonspecific and in selected cases FNAB may be helpful in patients with a history of extra-adrenal cancer, no other metastatic sites, and a heterogeneous adrenal mass with more than 20 HU, after exclusion of phaeochromocytoma (8), but only if the FNA information is likely to change management. When an extra-adrenal malignancy is not obvious, search for the primary tumour should be undertaken and total body scan and adrenal FNB are reasonable in this context. In only a limited number of cases an effective treatment of adrenal metastasis is available; however, the diagnosis of adrenal metastasis may change the therapeutic approach and has important implications on the clinical history of a cancer patient.

When evaluating the literature on radiological assessment of adrenal incidentalomas, we have to consider that almost all studies lack a definitive ascertainment of outcome since a pathological diagnosis based on the tumour specimen was available in a minority of cases. Final diagnosis was mainly based on the change in size of the adrenal mass over variable periods of observation.

Only two retrospective studies have evaluated ultrasonography in patients with adrenal incidentalomas and their outcomes were contradictory. The first study proposed ultrasonography as a first-line test in the follow-up of patients with adrenal incidentaloma, reporting that mass size was well correlated between ultrasonography and CT measurements, although ultrasonography could not differentiate mass type (18). The second study found that ultrasonography may detect only 65% of masses less than 3 cm (19).

CT is an accurate tool for detecting the presence of adrenal masses and differentiating between benign and malignant lesions. Using a fast scanner and 1-cm scanning intervals, both adrenals can be identified in 97–99% of patients. In previous studies, size has been reported to be the most reliable way to distinguish benign adenomas from ACCs, but more recent studies found that attenuation value is a superior parameter. Noncontrast CT attenuation coefficient expressed in Hounsfield units has been increasingly used to differentiate adrenal adenomas from nonadenomas (20). This is based on the fact that intracytoplasmic fat is often abundant in adrenal adenoma but is scarce in metastasis, phaeochromocytoma, or ACC (20). Threshold values for noncontrast CT ranging from 0 to 20 HU have been suggested and a cut-off value of 10 HU was recommended by a consensus panel organized by the National Institutes of Health (21). A density of 10 HU had the best accuracy, with a sensitivity of 96–100% and still a broad variability in specificity ranging from 50 to 100%. Adrenal masses with a density more than 10 HU on unenhanced CT required other tests for characterization (30% of adrenal adenomas have a low lipid content and may show higher attenuation values). Some studies suggested that lesions with density more than 43 HU on unenhanced CT should be considered as malignant (21).

In addition to lipid content, there have been a variety of other CT characteristics that may differentiate adrenal adenomas from nonadenomas. Such characteristics include smooth border, round or oval shape, sharp margins, maintenance of adrenal configuration, lack of calcification within or on the edge of the tumour, homogeneity of the mass, and lack of enhancement after contrast (22). Although these features are helpful in the characterization of a mass, none of them individually rules out malignancy with great confidence. Adrenal adenomas are often small, well-defined, homogeneous lesions that do not enhance on CT, and are believed to remain constant in size on serial CT scans (21). ACCs are usually large, dense, irregular, heterogeneous, enhancing lesions that invade other structures (21). However, small masses, in the range of 1 to 6 cm, may be difficult to discriminate (21, 22).

Enhanced CT is indicated when the mass density is more than 10 HU on unenhanced CT. An absolute washout at or above 40–60% 10 min after the administration of the contrast medium has a sensitivity of 82–96% and a specificity of 81–100% to differentiate benign from malignant masses. A relative washout of 37.5–50% 10 min after the administration of the contrast medium has sensitivity of 100% and specificity of 95–100% to differentiate benign from malignant masses (22).

MRI has also been used to differentiate between adrenal adenoma, metastasis, and phaeochromocytoma. Chemical shift MRI, similar to CT densitometry, is dependent on the detection of intracellular lipid in adenomas, but chemical shift MRI does so by relying on the different resonant frequencies of fat and water protons in a given voxel rather than on attenuation differences. Both T₁ and T₂ relaxation times have been studied. Signal intensity ratios between the adrenal mass and various organs, including spleen, fat, liver, and muscle, have been tested to discriminate adrenal masses. In general, malignant masses are denser than benign masses, though various benign lesions can mimic malignancies (21). The loss of signal on out-of-phase images in relation to spleen (to avoid the confounding of liver steatosis) differentiated adenomas from nonadenomas with a sensitivity of 84–100% and a specificity of 92–100% (21, 22). Studies suggest that there is no significant difference between CT and MRI for characterizing lipid-rich adenomas, whereas MRI might be superior when evaluating adrenal adenomas with a low lipid content with an attenuation value of up to 30 HU (22).

Two radioisotopes, ¹³¹I-iodomethyl-norcholesterol (NP-59) and 6-methyl-75-selenomethyl-19-norcholesterol have been used for imaging of adrenal masses of presumed adrenocortical origin. Various methods of analysing scintigraphy have been used to differentiate adrenal masses, including relative uptake of tracer, concordance with CT, and imaging patterns: adrenal nonadenomas have absent or significantly reduced uptake compared with adenomas (21). In addition to its potential role in diagnosing malignancy, scintigraphy may also be capable of differentiating autonomously secreting adenomas from nonfunctioning adenomas, adrenal hyperplasia, and other adrenal diseases (21). However, NP-59 adrenal scintigraphy is not reliable for lesions less than 2 cm in size. In recent years, the use of adrenal scintigraphy has declined because of the lack of widespread availability and parallel technical improvement of other radiological procedures (22).

[¹⁸F]2-fluoro-2-deoxy-d-glucose positron emission tomography (FDG PET) or PET/CT has also been reported to have a high sensitivity and specificity in characterizing adrenal lesions, although not as a routine imaging technique (22, 23), with a reported sensitivity of 93–100% and specificity of 80–100%. Necrotic or haemorrhagic malignant adrenal lesions may cause false-negative results showing poor FDG uptake. Metomidate-PET had a sensitivity of 89% and a specificity of 96% to differentiate masses of adrenal origin from masses of extra-adrenal origin. PET imaging is not reliable for lesions less than 1 cm in size, as metastatic lesions of this size may demonstrate less radiotracer uptake than normal liver. The use of PET/CT offers advantages over PET alone, as the morphology of the lesion can be assessed by CT while its metabolic activity can be measured concomitantly by PET, allowing for accurate anatomical localization of any focal FDG uptake. CT densitometry and washout measurements (if a delayed contrast-enhanced CT is performed) can be incorporated into the analysis. PET or PET/CT should be used when CT densitometry or washout analysis are inconclusive (22, 23).

Transcutaneous needle biopsy or FNAB, of an adrenal mass has been advocated by some for the investigation of incidentally discovered adrenal masses (21). FNAB is indicated only in patients with known extra-adrenal cancer when an adrenal adenoma has been reasonably excluded by CT or MRI (after biochemical exclusion of phaeochromocytoma). FNAB may be also useful in selected cases with discordant results of imaging tests and/or when rare tumours are suspected. The biopsy is generally performed under either CT or ultrasonography guidance. While accuracy appears to be high, up to 15% of biopsies are inconclusive (21). Complications of adrenal mass needle biopsy include pneumothorax, bleeding, and bacteraemia (21). Rare instances of metastatic seeding of the cancer along the needle track have been reported (21). A summary of recommendations for radiological assessment is outlined in Box 5.3.2.

With the exception of patients with imaging characteristics typical for myelolipoma or adrenal cyst, in all of the subjects with incidentally discovered adrenal mass either phaeochromocytoma or overt Cushing’s syndrome should be excluded (Box 5.3.3). Including patients with signs and symptoms attributable to an adrenal tumour that were overlooked before detection of an adrenal mass, will increase the proportion of secretory tumours. Conversely, using the strictest inclusion criteria and the purest definition of incidentaloma eliminates the need for considering overt Cushing’s syndrome.

As for ACC, there is little doubt that an early diagnosis of phaeochromocytoma is beneficial for the patient. Early recognition of the tumour may prevent potentially lethal hypertensive crises or arrhythmias (24). In the Italian survey, the frequency of phaeochromocytoma among adrenal incidentalomas was roughly comparable to that of adrenal carcinoma (approximately 4%) (6). Since an increasingly higher number of patients bearing a phaeochromocytoma are normotensive or have stable, low-grade hypertension, and phaeochromocytoma may not be easily recognized by imaging studies, it is mandatory to perform an appropriate biochemical screening in every patient with an adrenal mass according to the current guidelines (24).

Following recent epidemiological evidence that shows primary aldosteronism is the most frequent cause of endocrine hypertension, it was recommended to obtain a paired upright plasma aldosterone concentration and plasma renin activity in hypertensive patients with clinically inapparent adrenal adenoma in patients who are hypertensive. This measurement should be carried out after correction of hypokalaemia, if present. Dietary salt intake should be unrestricted. Hypokalaemia is no longer a mandatory prerequisite for suspecting primary hyperaldosteronism since more than 50% of patients are normokalaemic. Screening for primary aldosteronism should be pursued according to current guidelines (25).

In all patients with an incidentally discovered adrenal mass, the presence of overt cortisol excess must be suspected in the presence of one out the following four signs, which are relatively specific for Cushing’s syndrome: (1) easy bruising, (2) facial plethora, (3) proximal myopathy or muscle weakness, and (4) reddish-purple striae (>1 cm wide) (26). However, most patients with adrenal incidentalomas do not present signs or symptoms suggestive of hypercortisolism. If overt Cushing’s syndrome is not an issue, an endocrine work-up may frequently disclose subtle derangements of the hypothalamic–pituitary–adrenal axis (HPA) axis consistent with autonomous cortisol secretion by an incidental adrenal adenoma, the so-called subclinical Cushing’s syndrome (Box 5.3.4).

Subclinical Cushing’s syndrome may be defined as an autonomous cortisol secretion not fully restrained by pituitary feedback and variably exceeding the physiological daily production rate in the absence of an overt cushingoid phenotype (12, 27). Although the term ‘preclinical’ Cushing’s syndrome has been proposed previously, ‘subclinical’ Cushing’s syndrome more accurately describes this condition, not implying any assumption on the further development of a clinically overt syndrome (2). Since the prevalence of overt Cushing’s syndrome caused by adrenal adenoma in the general population is significantly lower than the prevalence of subclinical Cushing’s syndrome in patients with clinically nonfunctioning adrenal adenoma, it is rather inappropriate to consider subclinical Cushing’s syndrome as an early stage of development of overt hypercortisolism (27).

Although the pathophysiological concept of autonomous cortisol secretion sustained by an adrenal adenoma is straightforward, demonstration of subclinical Cushing’s syndrome is extremely difficult in practice. In fact, the standard biochemical tests used to screen Cushing’s syndrome are generally ill-suited to the assessment of patients who have no, or only mild, signs of cortisol excess. In this clinical setting, the a priori probability of subclinical Cushing’s syndrome is roughly comparable with the false-positive rate of the tests used for screening (2, 3). In the absence of reliable clinical clues it is indeed challenging to distinguish between true-positive and false-positive test results. Moreover, many tests used to study the HPA axis do not have sufficient sensitivity to recognize a very mild degree of cortisol excess. This is the case for the determination of urinary free cortisol, which has also the drawback of a remarkable daily variation in either cortisol excretion in the urine or daily urine output (the latter problem is amplified by the difficulty in obtaining complete urine collections) (27).

The reported prevalence of subclinical Cushing’s syndrome among patients with adrenal incidentaloma ranges from 5 to 20% (4, 8, 12, 27). This heterogeneity is explained, at least in part, by the different work-up protocols and variable criteria used to define subclinical cortisol excess as well as in different inclusion criteria and size of the reported series. Methodological limits add to the intrinsic biological problems associated with identification of subclinical cortisol excess, thus explaining the great uncertainty surrounding this entity. A number of alterations of the HPA axis have been associated to clinically inapparent adrenal adenomas and various biochemical criteria, alone or in combination, have been employed to qualify subclinical Cushing’s syndrome but the optimal diagnostic strategy remains to be defined (4, 8, 12, 27).

To provide a standard, in 2002, the National Institutes of Health state-of-the-science conference panel recommended the 1-mg dexamethasone suppression test as screening for subclinical Cushing’s syndrome with the traditional threshold of 5 μg/dl (138 nmol/l) to define adequate suppression (21). However, some experts advocate lower cut-points to increase detection of subclinical Cushing’s syndrome following the recommendations for screening of overt Cushing’s syndrome (4, 8). The rationale for this choice is that in most healthy subjects cortisol is barely detectable following 1 mg dexamethasone. However, specificity decreases when lower post-dexamethasone cortisol thresholds are used, which are likely to result in more false-positive test results (26). Conversely, other authors have suggested employing high-dose (3 or even 8 mg) dexamethasone tests since the diagnosis of pituitary Cushing’s syndrome is not a consideration (4, 8). At present, there is insufficient evidence to solve this controversy. However, the recommendation to use the overnight 1 mg suppression test seems sound since this test has been extensively employed for screening purposes, and the cut-off of 1.8 μg/dl (50 nmol/l) seems too low to assess individuals without specific features of hypercortisolism. The patients with an adrenal incidentaloma should be indistinguishable from the general population and in this setting the test specificity using this cut-off may be unacceptably low (12, 27).

Some experts require that two concomitant alterations in the tests aimed to study the HPA axis should be demonstrated to qualify a patient for subclinical Cushing’s syndrome, in order to circumvent the problem of false positivity of biochemical testing, and a number of tests have been employed for this purpose, thus making the screening procedure complicate and expensive (6). Blunting of the circadian rhythm of cortisol seems more frequent than elevation of urinary free cortisol and this confirms the view that derangement of the daily secretory pattern of cortisol is an early marker of (subclinical) hypercortisolism (27). Also, low to undetectable ACTH levels have been frequently reported, even if technical problems associated with measurement of ACTH concentrations close to the detection limits of the assay affect the utility of ACTH determination to demonstrate functional autonomy of an adrenal adenoma. Use of the corticotropin-releasing hormone test does not seem to add significant information to baseline ACTH levels (27). Recently, it has been demonstrated that the efficacy of midnight salivary cortisol in diagnosing subclinical Cushing’s syndrome is clearly lower than that found for overt Cushing’s syndrome (28).

The current uncertainty on what strategy is best suited to detect adrenal cortical autonomy might be solved by finding at what point cortisol excess becomes clinically significant, causing clinical morbidity. We are at present unable to answer this question because we do not know to what extent subclinical Cushing’s syndrome may affect patients’ health and life expectancy (12, 27).

Since many patients with clinically nonfunctioning incidentalomas are exposed to a chronic, even if only minimal to mild, cortisol excess, it is biologically plausible to anticipate that they should suffer, at least to some extent, from the classic, long-term consequences of overt Cushing’s syndrome, such as arterial hypertension, obesity, or diabetes (12, 21, 27). Several data from autopsy series, cross-sectional studies, and case–control studies (4, 8, 12, 17) consistently point to an association between clinically inapparent adrenal adenoma, subclinical Cushing’s syndrome, and the metabolic syndrome. There are also data suggesting that subclinical Cushing’s syndrome may predispose to osteoporosis, another well-established consequence of overt cortisol excess, and confers an increased risk of vertebral fractures (4, 12, 27). However, caution should be taken in generalizing results from series gathered in academic centres referral bias is an obvious issue since these studies are not population-based and there is the potential for confounding due to their case–control design. The complexity of an accurate matching between patients and controls for the many factors that may affect cardiovascular risk should also be disclosed. Moreover, the demonstration of an association should not imply a cause and effect relationship (27).

At present, subclinical Cushing’s syndrome presents a vexing problem as to diagnosis and management. The major areas of uncertainty are summarized in Box 5.3.5.

Management of adrenal incidentaloma is a complex decision-making process, which involves considering a range of possible diagnoses and their natural history, and weighing the risks and benefits of interventions in light of the patient’s age and the tumour size. Surgery is the appropriate therapeutic measure for ACC, phaeochromocytoma, and others functional adrenal tumours causing overt glucocorticoid, mineralocorticoid, or adrenal sex hormone; treatment of metastasis depends on individual clinical circumstances. Treatment of adrenal adenomas is much more difficult to outline because the natural history of these tumours is not well known (Box 5.3.6).

The available follow-up data of patients with clinically inapparent adrenal mass suggests that the large majority of adrenal lesions classified as benign at diagnosis remain stable over time. The risk of malignant transformation at long-term follow-up is very low, and it is estimated to be about 1:1000 incidentalomas (4). In 5–20% of cases mass size increases over time; however, most growing adrenal masses are not malignant (4, 27). The presence of isolated endocrine abnormalities at diagnosis may be considered a risk factor for mass enlargement or development of bilateral masses during follow-up (4). Occasional reduction or even disappearance of adrenal masses have been also reported in about 4% of adrenal incidentalomas, most often when cystic lesions, haematomas, or adrenal pseudotumours were the underlying diagnosis (12, 27).

The risk of progression from subclinical to overt Cushing’s syndrome is minimal (<1% of cases) (27). However, the occurrence of silent biochemical alterations during follow-up has been reported in a percentage ranging from 0 to 11% across different studies. The development of HPA axis abnormalities is unlikely in lesion smaller than 3 cm and appears to plateau after 3–4 years (4). A spontaneous regression of the alterations of the HPA axis may be observed, suggesting that cortisol output may have a cyclical or intermittent pattern (27).

The management of patients with subclinical Cushing’s syndrome is a very controversial issue. It is tempting to speculate that this condition represents a very mild variant of the syndrome of endogenous glucocorticoid excess sharing similar target-organ damages and long-term complications with the full-blown variant (12, 27). Evidence of increased morbidity and mortality in patients with clinically inapparent adrenal adenoma, with or without subclinical Cushing’s syndrome, is at present lacking and data are insufficient to indicate the superiority of a surgical or nonsurgical approach in the management of such patients (12, 27).

It is important to remember that patients with subclinical Cushing’s syndrome should receive perioperative glucocorticoids after removal of the functioning mass because they are at risk for hypoadrenalism. Factors such as young patient age, coexistence of hypertension, or diabetes, or osteoporosis might influence the decision in favour of surgery (4). The significant decrease in surgical morbidity and economic costs using a laparoscopic approach to adrenalectomy is actually widening indications to surgery (4). While adrenalectomy has been demonstrated to correct the HPA axis abnormalities, its effect on long-term patient outcome and quality of life is unknown. Until the risks and benefits of surgical removal of silent hyperfunctioning adrenocortical adenomas has been elucidated, we should elect to surgery patients with silent hypercortisolism who display diseases potentially attributable to cortisol excess that are of recent onset, or are resistant to medical intervention, or are rapidly worsening (12, 27). This strategy is based purely on pragmatism and not evidence. Box 5.3.7 outlines key issues with surgery in subclinical Cushing’s syndrome. Patients not submitted to surgery (possibly the majority) should undergo careful clinical monitoring and receive adequate treatment of the associated clinical conditions according to the specific guidelines (i.e. hypertension, diabetes) (27).

The limited and incomplete evidence available precludes making any stringent recommendation for periodic hormonal testing and repeat imaging evaluation for follow--up purposes. However, a repeat CT after 3 to 6 months from diagnosis should be recommended to recognize a rapidly growing mass whose malignant potential has escaped detection by the first imaging study, and then after 12 to 48 months (12). Hormonal testing (low-dose dexamethasone suppression test) is usually recommended in all patients with adrenal adenomas annually for 3–5 years. If no change in the functional state or imaging occurs further investigation may not be required (12, 27). However, it is important to stress the concept that little evidence is available to define the follow-up strategy, which should also consider the economic costs of follow-up investigations and the risk of cancer due to radiation exposure from multiple CT scans. Further research is urgently needed to inform a rational follow-up strategy, as outlined in Box 5.3.8.