doi: 10.1111/joim.12452
Genetics of adrenocortical tumours
. T. Åkerström1, T. Carling2, F. Beuschlein3 & P. Hellman1
From the 1 Department of Surgical Sciences, Uppsala University, Uppsala, Sweden; 2Endocrine Research Unit, Yale University, New Haven, CT, USA; and 3Medizinische Klinik und Poliklinik IV, Klinikum der Universität München, Munich, Germany
Content List - Read more articles from the symposium: Endocrine tumors - new generation sequencing with impact on therapy.
Abstract. Åkerström T, Carling T, Beuschlein F, Hellman P (Uppsala University, Uppsala, Sweden; Yale University, New Haven, CT, USA; Klinikum der Universität München, Munich, Germany). (Review Symposium). Genetics of adrenocortical tumours. J Intern Med 2016; 280: 540-550.
The recently available genomic sequencing tech- niques have led to breakthroughs in understand- ing of the underlying genetic mechanisms in adrenocortical tumours. Disease-causing muta- tions have been described for aldosterone-produ- cing adenomas, cortisol-producing adenomas and adrenocortical carcinomas. Further, knowledge gained from transcriptome analyses and methyla- tion arrays has provided new insights into the development of these tumours. Elucidation of the genomic landscape of adrenocortical tumours and improved techniques may in the future be useful
for early diagnosis through the detection of mutated DNA in the circulation. Moreover, com- pounds that bind specifically to altered proteins may be used as screening targets or therapeutic agents. Regulation of cortisol release by interaction with an altered subunit in adenylate cyclase may be more complex, but may provide a new option for regulating steroid release. Information about derangements in adrenocortical carcinoma is already helpful for determining patient prognosis. With further knowledge, we may be able to identify novel biomarkers that effectively and noninvasively help in differentiating between benign and malig- nant disease. It is clear that the next few years will provide much novel information that hopefully will aid in the treatment of patients with adrenocortical tumours.
Keywords: Adrenal tumours, genetics.
Introduction
During recent years, powerful genomic sequencing techniques have been readily available at a rea- sonable cost leading to breakthroughs in the genetic understanding of adrenal tumours. Although genetic derangements in adrenocortical carcinoma (ACC) have to some extent been previ- ously elucidated, the alterations in benign aldos- terone- and cortisol-producing adenomas (APAs and CPAs respectively) have until recently been unclear. In 2011, the first exomes of APAs were presented [1]. Following this discovery, next gen- eration techniques have helped elucidate genetic events in CPAs and ACCs. The identified mutated genes and altered proteins may be important targets for both new diagnostic methods, and treatment of general disorders such as hyperten- sion. Here, we summarize the current knowledge of genetic disturbances in adrenocortical tumours (Fig. 1).
Primary aldosteronism
Aldosterone is produced by cells located in the zona glomerulosa (ZG) of the adrenal cortex. It protects against hypovolaemia and salt wasting by increas- ing Na+ and H2O uptake by the kidney. The aldosterone-producing cells retain a high K+ per- meability and a membrane potential that approaches the equilibrium potential of K+ [2]. Ultimately, aldosterone production and release are regulated by changes in this electrical potential. Binding of angiotensin II to its receptor on the ZG cell surface and hyperkalaemia both cause depolarization of the cell, leading to increased intracellular Ca2+ and aldosterone secretion. Over- production of aldosterone without pathological consequence is observed in people living in warm and humid conditions with concomitant low salt intake, as exemplified by hill tribes in Papua New Guinea [3]. In countries in which salt intake is higher, hyperaldosteronism promotes hyperten-
CYP11B1/B2 germline mutations KCNJ5, CACNA1D, CACNA1H
somatic mutations KCNJ5, ATP1A1, ATP2B3, CACNA1D, CTNNB1
nodular hyperplasia
germline mutations PRKAR1A, PDE11A, PDE8B
TU
BAH
APA
somatic mutations CTNNB1
micronodular hyperplasia
TU
inactive adenoma
normal
adrenal ACTH
TU
somatic mutations PRKACA, GNAS
macronodular hyperplasia
germline and somatic ARMC5 mutations
CPA
TU
somatic mutations CTNNB1, TP53, CDKN2A, RB1 , ZNRF3 …
ACC
sion and cardiovascular damage, due to volume loading and direct actions on smooth muscle cells and fibroblasts. Primary aldosteronism (PA) is characterized by autonomous overproduction of aldosterone independent of its normal regulators. PA was previously considered to be rare, but with the clinical implementation of the aldosterone to renin ratio [4] the incidence of PA has risen and is now identified in 8-10% of all patients with essen- tial hypertension [5]. Two main subtypes of PA, bilateral idiopathic hyperplasia (IHA) and APAs constitute 95% of cases, whilst rare forms include familial disease and aldosterone-producing carci- nomas. Of note, unilateral disease (APA and uni- lateral/asymmetrical hyperplasia) can be treated surgically. As 30-50% of the adult population is hypertensive, PA represents a major health con- cern and needs to be properly diagnosed and treated.
Genetics of PA
Familial hyperaldosteronism
At present, different forms of familial hyperaldos- teronism (FH) can be stratified into subtypes based on germline genetic alterations [1, 6]. Familial hyperaldosteronism type I (FH I) is caused by fusion of the 5’ segment of the steroid 11ß- hydroxylase gene (CYP11B1) to the coding region of the aldosterone synthase enzyme (CYP11B2) [6]. Consequently, CYP11B2 becomes ectopically
expressed in zona fasciculata cells and regulated by adrenocorticotropic hormone (ACTH), explain- ing why glucocorticoids alleviate symptoms of the disease [6]. The genetic cause of the second disease type (FH II) remains to be established, although linkage in some families to chromosome 7p22 has been reported [7]. These patients, unlike FH I kindred, obtain no symptom relief from glucocor- ticoids and can appear to have sporadic disease, so that estimating disease prevalence can be chal- lenging. FH III was first described in a family with early onset hypertension, hybrid steroid produc- tion and no glucocorticoid response [8]. The genetic cause was found to be a germline mutation in the KCNJ5 gene, encoding an inwardly rectifying potassium channel (GIRK4) [1]. Following this discovery, multiple families with KCNJ5 mutations have been found. These patients have a diverse clinical phenotype that is dependent on the under- lying mutation and its effect on Na+ influx [9]. Mutations with relatively high levels of ion entry show no signs of hyperplasia on computed tomog- raphy scans, and usually mild disease that is controlled by medications, whereas lower influx leads to hyperplasia and a florid phenotype often requiring early bilateral adrenalectomy [9]. Of note, somatic KCNJ5 mutations are the most frequently altered gene in sporadic APAs [10, 11].
Recently, germline mutations in CACNA1D and CACNA1H have been observed in early onset PA
[12, 13]. These encode two a1-subunits in two different voltage-dependent calcium channels (VDCCs), which are responsible for the increase in Ca2+ concentration after depolarization of ZG cells [12, 13]. Both mutations lead to channel activation at lower membrane potentials and increased aldosterone production [12, 13]. Whilst patients with CACNA1D mutations develop a syndrome with PA, neuromuscular disease and seizures, CACNA1H kindred seem to develop only PA despite the gene being expressed in other organs [12, 13].
The genomic landscape of sporadic PA
The cause of IHA remains largely unknown. In a cohort of 251 patients with apparent sporadic disease, Murthy et al. [14] recently described three patients with germline KCNJ5 mutations (R52H, E246K and G247R) and nine patients with a rare nonsynonymous single nucleotide polymorphism (SNP; E282Q). These were all located outside the selectivity filter, and the R52H, E246K and E282Q polymorphisms were shown to have a functional consequence [14]. Of interest, a somatic E246G amino acid substitu- tion has been observed in one patient with a sporadic APA [15]. In addition, germline ARMC5 mutations have been observed in three patients with IHA [16]. The protein product contains multiple armadillo repeats, and is expressed in the ZG [16]. It was recently described as a tumour suppressor in corticotropin-independent macronodular hyperplasia [17]. Although its role in PA is still not fully understood, animal models are being developed and may provide additional information about the consequence of these alterations.
The genetic understanding of APAs, in contrast to IHA, has been revolutionized in the last few years. The first exome sequencing on APAs revealed a low number of somatic events per megabase, and remarkably in 2/4 of the analysed tumours a somatic mutation in the KCNJ5 gene, subsequently verified worldwide in over 1000 tumours [10, 11, 18-22]. The normal function of GIRK4 in ZG cells is to lower aldosterone production by promoting K+ efflux (Fig. 2) [23]. These channels contain a highly conserved TXGYGFR motif, constituting the chan- nel selectivity filter [24]. The majority of KCNJ5 mutations disrupt this filter leading to loss of selectivity for K+, increased inward Na+ current, depolarization and enhanced aldosterone secretion
[1, 25]. Patients with tumours carrying KCNJ5 mutations have a distinct phenotype compared to those with wild-type KCNJ5 APAs [26]. Patients with these mutations have higher aldosterone levels [10], undergo surgery at a younger age [10, 15, 27], are more likely to be female [10, 11, 15, 28] and have larger tumours [11, 15, 18]. The female overrepresentation is intriguing, and may reflect differences in the clinical presentation due to protective effects of testosterone or deleterious effects of oestradiol.
In vitro, mutations in KCNJ5 have not shown any effect on proliferation, raising the possibility that additional mechanisms may be involved. Influx of positive ions promotes cell proliferation, but can also increase cell death at higher levels, illustrated in kindreds with germline KCNJ5 mutations where hyperplasia and severity of the disease depend upon the level of depolarization. For an APA to develop, there is likely to be a balance between protective, proliferative and damaging effects. These changes may be difficult to study with cell lines due to their complex nature, although some interesting results have been reported [29]. Find- ings from studies in kindreds with bilateral hyper- plasia with germline KCNJ5 mutations suggest that this mutation is sufficient to cause both increased aldosterone production and proliferation in these patients at least [1].
Analyses of large multicentre cohorts have estab- lished that the prevalence of KCNJ5 mutations in western countries is approximately 40% [10, 11]. In Asian countries the prevalence seems to be higher, with at least two-thirds of all APAs carry- ing KCNJ5 mutations [21]. The reasons for this discrepancy remain unknown but may include demographic differences such as salt intake [26] or variations in clinical routine. Asian cohorts also seem to have a lower prevalence of ATP1A1, ATP2B3 and CACNA1D mutations, suggesting that environmental effects promote KCNJ5 muta- tions or selectively worsen the phenotype in these patients. Of note, it has recently been shown that the use of adrenal vein sampling may affect the prevalence estimates of different mutations, prob- ably due to increased diagnosis of smaller tumours [19].
The transcriptomic profiles of APAs with KCNJ5 mutations have been investigated in different cohorts. In a large profiling study of 102 APAs, no specific mRNA signatures in tumours with
(a)
Normal cell
(b) Normal cell - AT1R stimulation
(c)
Tumour cell
K *- channels (GIRK, TASK etc.)
K *- channels (GIRK, TASK etc.)
Na+/K *- ATPase
Mutated K+-channel (GIRK4)
Na+/K *- ATPase
Mutated Na+/K+-ATPase Na+ H+
+
AT1R
K
+
AT1R
Angli
K
!
AT1R
1
Na+
Na+
K*
Voltage-gated Ca2+ channel
K*
Voltage-gated Ca2+ channel Ca2+
+ Na+
Mutated Voltage-gated Ca2+ channel Ca2+
Hyperpolarization
Depolarization
Depolarization
Ca2+ ATPase Na+
[Ca2+]
Ca2+ ATPase Na*
[Ca2+]
NUCLEUS
NUCLEUS
Mutated Ca2+ ATPase Cations
NUCLEUS
Aldosterone
Aldosterone
CYP11B2
CYP11B2
CYP11B2
KCNJ5 mutations were identified [10]. However, the results from three smaller studies suggested distinct signatures in mutated tumours [30-32]. Interestingly, increasing evidence also suggests that tumours with KCNJ5 mutations have a zona fasciculata like cell constitution, relatively high expression of CYP11B1 and low expression of CYP11B2 [31-34]. This indicates that different genetic aberrations might lead to subtle pheno- typic variations, or that APAs originate from different cell types. Other genetic aberrations recently identified in APAs are mutations in the ion channel CACNA1D and in two ATPases (ATP1A1 and ATP2B3) [12, 31, 35]. Like GIRK4, these are of importance for regulating the mem- brane potential and Ca2+ homoeostasis in ZG cells (Fig. 2b). However, in contrast to tumours with KCNJ5 mutations, these seem to have a ZG-like cell constitution and higher CYP11B2 expression [31, 33, 34]. Remarkably, somatic mutations in ATP1A1, ATP2B3 and CACNA1D were recently described in aldosterone-producing cell clusters from normal adrenal glands, suggesting that these constitute a precursor to APAs, and perhaps may predispose for ‘subclinical’ hyperaldostero- nism associated with increased morbidity. Inter- estingly, no mutations in KCNJ5 has yet been detected in these clusters, suggesting either that these mutations rapidly lead to APA formation and hence are not detected [36] or perhaps that mutated tumours are indeed derived from differ- ent precursors.
CACNA1D encodes the a1-subunit of a VDCC that increases Ca2+ influx in response to depolarization (Fig. 2). It is highly expressed in the ZG and its activation stimulates aldosterone synthesis [2]. The a1-subunit forms the channel pore and is composed of four repeated domains, each contain- ing six transmembrane segments. Mutations have been found in hotspot areas, most accumulating in segment 4 (S4), S5 and S6 in each repeat [12, 19, 31, 32]. The S5 and S6 segments line the channel pore and most mutations occur in these segments. In vitro, the mutations cause channel activation at more hyperpolarized levels, increasing intracellular Ca2+ and aldosterone production, and phenocopy- ing the effect of KCNJ5 mutants [12, 31]. Despite this, patients with CACNA1D mutations seem to have a different clinical phenotype, associated with smaller adenomas, male gender and older age at surgery [19, 31, 32].
ATP1A1 encodes the a1-subunit of a ubiquitously expressed Na+/K+-ATPase [37]. At rest, this protein establishes and maintains the negative cell mem- brane potential by exchanging two extracellular K+ ions for three cytoplasmic Na+ ions whilst consum- ing ATP (Fig. 2). Somatic mutations in ATP1A1 are found in about 5% of APAs [19, 31, 32, 35]. Mutations are located in the M1, M4 and M9 domains of the protein. The M1 and M4 domains are juxtaposed, and the residue Leu104 interacts with Glu334 to form a K+ ion-binding site that is vital for protein function. The mutations in M1 are
located close to or involve this Leu104 residue, whilst mutations in the M4 domain alter the Val332 residue. It is likely that both mutations distort the ion-binding pocket. In vitro, mutants show lower K+ affinity and increased depolarization [31, 35], probably due to increased influx of positively charged ions [31]. It has been hypothe- sized that the M9 domain contains the third Na+- binding site of the protein, and the Glu961 amino acid is necessary for proper function [38]. Of note, an inward proton current has been demonstrated at this site in the nonmutated Na+/K+-ATPase, especially during delayed K+ ion binding and occlusion [39], suggesting that mutations at this site might allow for increased cationic influx. Genotype-phenotype correlation studies have shown that mutations in ATP1A1, like CACNA1D mutants, are associated with male gender [19, 31] and smaller tumours [31].
Another ATPase, ATP2B3, is mutated in 1.5% of APAs [19, 20, 32, 35]. The protein product is a plasma membrane Ca2+-ATPase (PMCA3) that utilizes ATP to transport Ca2+ from the cytosol [40]. Similar to the Na+/K+-ATPase x1-subunit structure, the PMCA3 is composed of 10 trans- membrane segments, and all hitherto identified mutations are located in the M4 domain of the protein [19, 20, 35]. These affect a valine residue (Val426) homologous to the valine mutated in the M4 domain in ATP1A1 (p.Val332). These deletions are likely to distort the Ca2+ ion-binding motifs leading to increased depolarization and enhanced aldosterone production [35].
The WNT signalling pathway
In the adrenal cortex, active WNT signalling is restricted to ZG cells [41]. Disrupting the WNT pathway results in disturbances of adrenal devel- opment and reduced aldosterone levels [42]. Of interest, mice with activating mutations in the ß- catenin gene (CTNNB1) develop PA and tumours that can become malignant [43]. In human APAs, two-thirds of tumours exhibit active WNT sig- nalling [41]. Decreased expression of the negative WNT regulator secreted frizzled-related protein II (SFRP2) may explain this activation in some tumours [44]. Mutations in CTNNB1 have also been identified in APAs using exome sequencing [12, 31]. However, no mutations were detected in a series of 41 APAs [41]. Our own unpublished results show a prevalence of 5% mutations, suggesting that these do occur, albeit at a
relatively low prevalence (Åkerström et al. sub- mitted). By contrast, CTNNB1 mutations have been observed at a higher frequency in other adrenocortical tumours as well as in ACCs [45, 46]. Notably, young mice with mutations in CTNNB1 present with hyperaldosteronism. How- ever, there is a trend towards lower aldosterone levels and increased corticosterone production as the mice age [43], indicating a possible dediffer- entiation of the aldosterone-producing cells, which might explain a lower prevalence of mutations in APAs.
An intriguing genetic landscape in PA has been revealed through the implementation of new molec- ular techniques. Remarkably, ~70% of APAs have been found to harbour a pathological mutation. These genetic discoveries have improved not only knowledge of PA, but also understanding of the normal physiology of ZG cells and the effects of ion homoeostasis on cell proliferation and tumourige- nesis in general. Future challenges include increasing the awareness and screening of PA in hypertensive individuals, and adopting less inva- sive techniques for detection of surgically resect- able disease. With increased knowledge of the molecular pathways affected, new approaches might be found to overcome these challenges.
Glucocorticoid-producing adenomas
Human cortisol production is tightly regulated. Stress signals cause release of corticotropin-releas- ing hormone from the hypothalamus, leading to pituitary secretion of ACTH. In the adrenal fascic- ulata, ACTH binds to the G-protein-coupled recep- tor melanocortin receptor 2, which induces GTP binding to Ga. The GTP-bound Ga activates adeny- late cyclase followed by cAMP binding to the regulatory subunit of protein kinase A (PKA), which induces release of the PKA catalytic subunit and enables downstream signalling. Cortisol biosyn- thesis is acutely enhanced through conversion of cholesterol to pregnenolone. It has been demon- strated that this process involves PKA phosphory- lation of cAMP response element-binding protein (CREB) and other proteins. Chronically this leads to increased expression of all enzymes of the cortisol biosynthetic pathway, and enhanced adrenocortical mass via increased cell prolifera- tion. In the normal physiological environment, increased cortisol levels provide negative feedback, inhibiting further production of ACTH at both the hypothalamic and pituitary levels [47].
Classic Cushing’s syndrome is characterized by the signs and symptoms of hypercortisolism such as moon face, buffalo hump, central obesity, easy bruising, deep purple striae, acne, proximal mus- cle wasting, hirsutism and glucose intolerance. The syndrome is rare with an overall annual incidence of 1 in 50 000 [48]. Primary adrenal lesions account for about 15-25% of cases of Cushing’s syndrome. However, in most studies a prevalence of 5-24% has been demonstrated for subclinical hypercortisolism in patients with an adrenal inci- dentaloma [49]. This broad range may be attrib- uted to the different diagnostic criteria used over time.
Molecular genetics of glucocorticoid-producing adenomas
Evidence had previously suggested that CPAs, like APAs, exhibit a monoclonal lesion with varying degrees of chromosomal loss or gain [50, 51]. Moreover, rare germline or mosaic mutations caus- ing increased Go activity can also cause Cushing’s syndrome, typically due to bilateral adrenal hyper- plasia. For example, germline mosaicism for activating mutations in GNAS leads to McCune- Albright syndrome, which may include Cushing’s syndrome due to adrenal hyperplasia [52]. Simi- larly, germline and somatic loss-of-function mutations in PRKAR1A (encoding the regulatory subunit of PKA) have been found in up to 80% of patients with Carney complex, which includes Cushing’s syndrome due to primary pigmented nodular adrenocortical disease (PPNAD) [53, 54]. In addition, sporadic adrenal adenomas (typically with a PPNAD-like clinical presentation of Cush- ing’s syndrome) also result from somatic mutation of PRKAR1A [54].
Recently, exome sequencing was employed to identify somatic mutations involved in CPAs by four independent research groups [55-58]. Somatic recurring activating mutations of the main catalytic subunit of PKA, PRKACA (c.617A > C, also known as c.617T > G), were discovered result- ing in arginine substitution of amino acid 206 (Leu206Arg) in unilateral CPAs with overt Cush- ing’s syndrome (Fig. 3). Beuschlein et al. [55] pre- sented results from 139 patients with adrenal adenomas, ACCs and ACTH-independent primary adrenal hyperplasia. The authors identified muta- tions in PRKACA in eight of the 10 originally screened unilateral CPAs; seven of the patients demonstrated the Leu206Arg mutation, whereas one had the insertion located at Leu199_Cy-
s200insTrp. The Leu206Arg mutation is located in a highly conserved core of the interaction between the regulatory (RIIß) and catalytic sub- units of PKA. Goh et al. [57] reported the results of exome sequencing of cortisol-producing tumours from 25 patients (22 with adrenocortical adenomas and three with ACCs). They demonstrated the identical heterozygous PRKACA somatic mutation Leu206Arg in six patients. The described hotspot mutations may interfere with the creation of a stable PKA molecule and render the mutant Cơ subunits constitutively active. Beuschlein et al. described the clinical genotype-phenotype associ- ations in these patients: 37% of the studied cohort had overt Cushing’s syndrome due to a unilateral CPA with a PRKACA mutation. These patients had a higher index of disease severity, as shown by increased urinary free cortisol and late-night serum cortisol levels. These findings were corre- lated with expression levels of the steroidogenic enzymes that were higher in tissues with PRKACA mutations [55]. Similarly, Goh et al. demonstrated an increased expression of downstream target genes consistent with enhanced steroidogenic enzymatic activity in the adrenal tissue with PRKACA mutations, causing tumour development and endogenous Cushing’s syndrome. Moreover, patients with mutant adrenal adenomas were younger and had smaller tumours, associated with overt Cushing’s syndrome rather than subclinical hypercortisolism.
Independently, the other two groups published their finding in the same issue of the journal Science. Cao et al. [56] analysed 49 samples, including CPAs, carcinomas and ACTH-indepen- dent macronodular adrenal hyperplasia tumours using whole-exome and RNA sequencing. The authors demonstrated an even higher rate of the somatic Leu206Arg PRKACA mutation (69.2%). Of note, no statistically significant differences in serum cortisol, plasma ACTH and urinary free cortisol levels were found compared with patients with wild-type PRKACA. Functional overexpression of Leu205Arg mutants (gain-of-function mutation in 293T cells) led to higher PKA-target phosphory- lation compared to wild-type. In agreement with the findings of other studies, the PRKACA Leu206Arg mutation induced phosphorylation of CREB, confirming the hypothesis that this muta- tion may enhance PKA activity, causing tumour development and endogenous Cushing’s syn- drome. Similarly, Sato et al. performed whole- exome sequencing on eight adrenal CPAs, and
(a)
Normal cell
(b) Normal cell - ACTH stimulation
(c) Tumour cell - PRKACA mutation
Mineralocorticoid receptor 2
Mineralocorticoid receptor 2
Mineralocorticoid receptor 2
Adenylate cyclase
ACTH
Adenylate cyclase
CAMA
Adenylate cyclase
AC
By
By
AC
Go,
AC
Ga
PKAR
PKAR
PKAR
CAMP
Inactive PKAc
Active PKAc
Mutated PKAc
NUCLEUS
NUCLEUS
NUCLEUS
CREB - P
Cortisol
CREB - P
Cortisol
Steroidogenic enzymes
Steroidogenic enzymes
found that four displayed the PRKACA Leu206Arg mutation. Further analysis demonstrated the iden- tical mutation in 24/57 tumours. Similar to the findings of Goh et al., the affected patients had significantly smaller tumours and more pro- nounced Cushing’s syndrome as evidenced by higher levels of serum cortisol after the 1 mg dexamethasone suppression test [58].
Recently, Di Dalmazi et al. demonstrated that the PRKACA Leu206Arg missense mutation was pre- sent in 34% of 149 clinical cases, and was clinically associated with Cushing’s syndrome. In addition to this previously described mutation, the authors identified two novel mutations in another four patients (c.600_601insGTG/p.Cys200_Gly201ins- Val and c.639C>G+c.638_640insATTATCCTGAGG /p.Ser213Arg+p.Leu212_Lys214insIle-Ile-Leu- Arg) [59].
The identification of Leu206Arg PRKACA muta- tions in a high proportion of CPAs provides an example of the genotype-phenotype correlations often observed in endocrine neoplasms. The corre- lation might reflect the highly differentiated state of many endocrine tumours and their genomic simplicity and low overall mutation densities.
Adrenocortical carcinoma
Adrenocortical carcinoma is a rare tumour that affects both adults and children with an annual
incidence of 0.7-2.0 cases per million individuals. The disease is highly malignant with only 16-38% of patients surviving more than 5 years after initial diagnosis [60-63]. According to their cellular origin, adrenocortical cancer cells maintain steroidogenic properties with biochemical studies showing that 60-70% of patients overproduce hormones. However, in contrast to benign hormonally active adrenocortical adenomas, steroid production in ACCs is not clinically apparent in all cases due to relatively inefficient production characterized by increased levels of steroid precursors. Indeed, this dedifferentiated and thus incomplete pattern of steroidogenic enzyme expression can be clearly revealed by steroid profiling, providing a malignant steroid fingerprint in ACC patients [64].
The only drug approved for adjuvant ACC treatment as well as for advanced tumour stages is mitotane. This agent, which exerts inhibitory effects on both steroidogenesis and on cell growth, should prefer- ably be used in combination with etoposide, dox- orubicin and cisplatin [65]. However, even in the setting of a clinical study, systemic treatments have achieved only partial responses and even after radical resection as many as 85% of patients relapse [62]. Therefore, a number of studies have been conducted in order to identify biomarkers with diagnostic and prognostic impact to improve initial diagnosis as well as follow-up of patients with ACC. Furthermore, the advent of novel powerful omic techniques has provided hope that the elucidation of
molecular mechanisms could translate into targeted therapeutic strategies.
ACC as part of a genetic syndrome
From the time of its initial description in 1967, ACC has been associated with Li-Fraumeni syndrome [66]; it was later recognized that this syndrome is caused by germline TP53 mutations. Furthermore, TP53 mutations were found in a high proportion of paediatric ACC patients in several cohorts. In areas with a particularly high incidence of childhood ACC, such as southern Brazil, newborn screening has been able to identify a low penetrance TP53 mutation in around 0.3% of cases [67]. Although amongst those patients under surveillance only a small fraction of gene carriers developed ACCs, tumour size and stage were low providing evidence for the effectiveness of this screening procedure.
Adrenocortical hyperplasia and tumours of vari- able malignancy are found in cases with the rare Beckwith-Wiedemann syndrome, a systemic over- growth disorder caused by genetic defects that ultimately result in overexpression of insulin-like growth factor 2 (IGF2) [68]. In fact, loss of heterozy- gosity of the 11p15 region, which harbours the IGF2 locus, is a common finding in childhood ACC [69], and is associated with increased expression of IGF2 [70]. This molecular feature thereby mimics the situation in the foetal adrenal where IGF2 has an important role in regulating proliferation and steroid production. Interestingly, also in adult ACCs, IGF2 overexpression and abnormalities in the 11p15 region is amongst those molecular alterations with the most distinct differences to benign adrenal tumours [71, 72]. By contrast, it is clear from mouse models overexpressing Igf2 that this growth factor can induce adrenal hyperplasia but not malignant tumourigenesis per se [73, 74]. Similarly, a recently published clinical trial of an oral small molecule inhibitor of both IGF1 receptor and the insulin receptor in patients with advanced ACC demonstrated no benefit in terms of overall survival with only a few patients experience long- term stable disease [75].
Genomic landscape of ACCs
There is increasing evidence of an association between adult ACC and several inherited condi- tions, such as the above mentioned Li-Fraumeni syndrome [76], and multiple endocrine neoplasia type 1 [77]. In addition, it has been recognized that
ACC is a malignancy associated with Lynch syn- drome, an inherited disorder that predisposed towards an increased risk of certain cancers, including colorectal and endometrial cancers, due to heterozygous mutations in mismatch repair genes [78]. However, the majority of adult-onset ACC cases can be considered to occur as sporadic disease, which is defined by the acquisition of somatic mutations.
The use of different genetic and molecular method- ologies had already revealed a clear distinction between malignant and benign adrenocortical lesions with regard to chromosomal aberrations [79], transcriptomes (mRNA and microRNA expres- sion profiles) [80] and epigenomic characteristics (DNA methylation profiles) [81]. Furthermore, groups of patients could be identified with distinct clinical features amongst ACC cases. For example, genomewide methylation analysis revealed the existence of a cohort with hypermethylated tumours with a poorer prognosis, providing indi- rect evidence that hypermethylation is an impor- tant mechanism for silencing specific tumour suppressor genes [81]. Similarly, unsupervised clustering of transcriptome data demonstrated different subgroups, which were associated with p53 or ß-catenin alterations [80], and had defined markers of good and poor prognosis [82]. At the genetic level, combined exome sequencing and SNP analysis revealed recurrent alterations in known driver genes (CTNNB1, TP53, CDKN2A, RB1 and MEN1) as well as in genes not previously reported in ACC (ZNRF3, DAXX, TERT and MED12) [46].
The elucidation of genomic alterations in ACCs has recently gained further momentum with the inte- grative data analysis of genetic and epigenetic alterations and gene expression profiles [83]. Indeed, substantial overlap between these classifi- cations has been identify through the integration of different omics approaches, with results from tran- scriptome studies strongly correlating with DNA methylation profiles and mutation rates associated with specific expression pattern [46]. These findings provide the basis for the molecular classification of adult ACC with prognostic value beyond that of histological markers currently in clinical use [84]. Similarly, the genomic landscape of childhood ACC has been elucidated [85]. In addition to the known germline alterations and IGF2 overexpression as described above, further genetic alterations includ- ing recurrent somatic mutations in ATRX and CTNNB1 and integration of human herpes virus 6
in chromosome 11p have been identified. Thus, although some genetic events are common for adult and paediatric ACCs, there are specific features that distinguish the tumours according to their occur- rence early or late in life.
Collectively, these discoveries have created the opportunity for classifying adrenocortical tumours on the basis of molecular analyses. Following these genomic studies, there have been efforts to develop new molecular tools to improve diagnosis and prog- nostication in patients with adrenocortical tumours in clinical practice. It will be informative to deter- mine molecular phenotype correlations in ACC patients that could relate to steroid profiles, which would allow for noninvasive approaches for diagno- sis and follow-up. At present, it is clear that the genetic make-up of benign and malignant adreno- cortical tumours differs not only according to the numbers of genetic events but also with regard to the genes affected by germline and/or somatic muta- tions. However, CTNNB1 mutations occur in both benign and malignant disease, and some clinical aspects are shared between benign and malignant adrenal tumours, such as the occurrence of adrenal Cushing’s syndrome and development of an adrenal mass. If an adenoma to carcinoma sequence is present, it is likely exceedingly rare.
Conflict of interest
No conflicts of interest to declare.
References
1 Choi M, Scholl UI, Yue P et al. K+ channel mutations in adrenal aldosterone-producing adenomas and hereditary hypertension. Science 2011; 331: 768-72.
2 Spät A, Hunyady L. Control of aldosterone secretion: a model for convergence in cellular signaling pathways. Physiol Rev 2004; 84: 489-539.
3 Rossi GP, Boscaro M, Ronconi V, Funder JW. Aldosterone as a cardiovascular risk factor. Trends Endocrinol Metab 2005; 16: 104-7.
4 Hiramatsu KYT, Yukimura Y, Komiya I et al. A screening test to identify aldosterone-producing adenoma by measuring plasma renin activity: results in hypertensive patients. Arch Intern Med 1981; 141: 1589-93.
5 Rossi GP, Bernini G, Caliumi C et al. A prospective study of the prevalence of primary aldosteronism in 1,125 hyperten- sive patients. J Am Coll Cardiol 2006; 48: 2293-300.
6 Lifton RP, Dluhy RG, Powers M et al. A chimaeric 11 beta- hydroxylase/aldosterone synthase gene causes glucocorti- coid-remediable aldosteronism and human hypertension. Nature 1992; 355: 262-5.
7 Lafferty AR, Torpy DJ, Stowasser M et al. A novel genetic locus for low renin hypertension: familial hyperaldosteronism type II maps to chromosome 7 (7p22). J Med Genet 2000; 37: 831-5.
8 Geller DS, Zhang J, Wisgerhof MV, Shackleton C, Kashgarian M, Lifton RP. A novel form of human mendelian hypertension featuring nonglucocorticoid-remediable aldosteronism. J Clin Endocrinol Metab 2008; 93: 3117-23.
9 Scholl UI, Nelson-Williams C, Yue P et al. Hypertension with or without adrenal hyperplasia due to different inherited mutations in the potassium channel KCNJ5. Proc Natl Acad Sci USA 2012; 109: 2533-8.
10 Boulkroun S, Beuschlein F, Rossi GP et al. Prevalence, clinical, and molecular correlates of KCNJ5 mutations in primary aldosteronism. Hypertension 2012; 59: 592-8.
11 Akerstrom T, Crona J, Delgado Verdugo A et al. Comprehen- sive re-sequencing of adrenal aldosterone producing lesions reveal three somatic mutations near the KCNJ5 potassium channel selectivity filter. PLoS One 2012; 7: e41926.
12 Scholl UI, Goh G, Stolting G et al. Somatic and germline CACNA1D calcium channel mutations in aldosterone-produ- cing adenomas and primary aldosteronism. Nat Genet 2013; 45: 1050-4.
13 Scholl UI, Stölting G, Nelson-Williams C et al. Recurrent gain of function mutation in calcium channel CACNA1H causes early-onset hypertension with primary aldosteronism. Elife 2015; 4: e06315.
14 Murthy M, Xu S, Massimo G et al. Role for germline mutations and a rare coding single nucleotide polymorphism within the KCNJ5 potassium channel in a large cohort of sporadic cases of primary aldosteronism. Hypertension 2014; 63: 783-9.
15 Cheng CJ, Sung CC, Wu ST et al. Novel KCNJ5 mutations in sporadic aldosterone-producing adenoma reduce Kir3.4 membrane abundance. J Clin Endocrinol Metab 2015; 100: E155-63.
16 Zilbermint M, Xekouki P, Faucz FR et al. Primary aldostero- nism and ARMC5 variants. J Clin Endocrinol Metab 2015; 100: E900-9.
17 Assie G, Libe R, Espiard S et al. ARMC5 mutations in macronodular adrenal hyperplasia with Cushing’s syndrome. N Engl J Med 2013; 369: 2105-14.
18 Azizan EA, Murthy M, Stowasser M et al. Somatic mutations affecting the selectivity filter of KCNJ5 are frequent in 2 large unselected collections of adrenal aldosteronomas. Hyperten- sion 2012; 59: 587-91.
19 Fernandes-Rosa FL, Williams TA, Riester A et al. Genetic spectrum and clinical correlates of somatic mutations in aldosterone-producing adenoma. Hypertension 2014; 64: 354-61.
20 Dutta RK, Welander J, Brauckhoff M et al. Complementary somatic mutations of KCNJ5, ATP1A1 and ATP2B3 in spo- radic aldosterone producing adrenal adenomas. Endocr Relat Cancer 2013; 21: L1-4.
21 Taguchi R, Yamada M, Nakajima Y et al. Expression and mutations of KCNJ5 mRNA in Japanese patients with aldos- terone-producing adenomas. J Clin Endocrinol Metab 2012; 97: 1311-9.
22 Zheng FF, Zhu LM, Nie AF et al. Clinical characteristics of somatic mutations in Chinese patients with aldosterone- producing adenoma. Hypertension 2015; 65: 622-8.
23 Oki K, Plonczynski MW, Lam ML, Gomez-Sanchez EP, Gomez- Sanchez CE. The potassium channel, Kir3.4 participates in
angiotensin II-stimulated aldosterone production by a human adrenocortical cell line. Endocrinology 2012; 153: 4328- 35.
24 Tao X, Avalos JL, Chen J, Mackinnon R. Crystal structure of the eukaryotic strong inward-rectifier K+ channel Kir2.2 at 3.1 A resolution. Science 2009; 326: 1668-74.
25 Oki K, Plonczynski MW, Luis Lam M, Gomez-Sanchez EP, Gomez-Sanchez CE. Potassium channel mutant KCNJ5 T158A expression in HAC-15 cells increases aldosterone synthesis. Endocrinology 2012; 153: 1774-82.
26 Lenzini L, Rossitto G, Maiolino G, Letizia C, Funder JW, Rossi GP. A meta-analysis of somatic KCNJ5 K channel mutations in 1636 patients with an aldosterone-producing adenoma. J Clin Endocrinol Metab 2015; 100: E1089-95.
27 Kitamoto T, Suematsu S, Matsuzawa Y, Saito J, Omura M, Nishikawa T. Comparison of cardiovascular complications in patients with and without KCNJ5 gene mutations harboring aldosterone-producing adenomas. J Atheroscler Thromb 2015; 22: 191-200.
28 Kuppusamy M, Caroccia B, Stindl J et al. A novel KCNJ5- insT149 somatic mutation close to, but outside, the selec- tivity filter causes resistant hypertension by loss of selectiv- ity for potassium. J Clin Endocrinol Metab 2014; 99: E1765- 73.
29 Williams TA, Monticone S, Crudo V, Warth R, Veglio F, Mulatero P. Visinin-like 1 is upregulated in aldosterone- producing adenomas with KCNJ5 mutations and protects from calcium-induced apoptosis. Hypertension 2012; 59: 833-9.
30 Monticone S, Hattangady NG, Nishimoto K et al. Effect of KCNJ5 mutations on gene expression in aldosterone-produ- cing adenomas and adrenocortical cells. J Clin Endocrinol Metab 2012; 97: E1567-72.
31 Azizan EA, Poulsen H, Tuluc P et al. Somatic mutations in ATP1A1 and CACNA1D underlie a common subtype of adrenal hypertension. Nat Genet 2013; 45: 1055-60.
32 Åkerström T, Willenberg HS, Cupisti K et al. Novel somatic mutations and distinct molecular signature in aldosterone- producing adenomas. Endocr Relat Cancer 2015; 5: 735- 44.
33 Dekkers T, Ter Meer M, Lenders JW et al. Adrenal nodularity and somatic mutations in primary aldosteronism: one node is the culprit? J Clin Endocrinol Metab 2014; 99: E1341- 51.
34 Monticone S, Castellano I, Versace K et al. Immunohisto- chemical, genetic and clinical characterization of sporadic aldosterone-producing adenomas. Mol Cell Endocrinol 2015; 411: 146-54.
35 Beuschlein F, Boulkroun S, Osswald A et al. Somatic muta- tions in ATP1A1 and ATP2B3 lead to aldosterone-producing adenomas and secondary hypertension. Nat Genet 2013; 45: 440-4.
36 Nishimoto K, Tomlins SA, Kuick R et al. Aldosterone-stimu- lating somatic gene mutations are common in normal adrenal glands. Proc Natl Acad Sci USA 2015; 112: E4591-9.
37 Jorgensen Peter L, Håkansson KO, Karlish SJD. Structure and mechanism of Na, K-ATPase: functional sites and their interactions. Annu Rev Physiol 2003; 65: 817-49.
38 Li C, Capendeguy O, Geering K, Horisberger JD. A third Na+- binding site in the sodium pump. Proc Natl Acad Sci USA 2005; 102: 12706-11.
39 Vedovato N, Gadsby DC. Route, mechanism, and implications of proton import during Na+/K+ exchange by native Na+/K+- ATPase pumps. J Gen Physiol 2014; 143: 449-64.
40 Schatzmann HJ. ATP-dependent Ca++-extrusion from human red cells. Experientia 1966; 22: 364-5.
41 Boulkroun S, Samson-Couterie B, Golib-Dzib JF et al. Aldos- terone-producing adenoma formation in the adrenal cortex involves expression of stem/progenitor cell markers. Endocrinology 2011; 152: 4753-63.
42 Kim AC, Reuter AL, Zubair M et al. Targeted disruption of beta-catenin in Sf1-expressing cells impairs development and maintenance of the adrenal cortex. Development 2008; 135: 2593-602.
43 Berthon A, Sahut-Barnola I, Lambert-Langlais S et al. Con- stitutive beta-catenin activation induces adrenal hyperplasia and promotes adrenal cancer development. Hum Mol Genet 2010; 19: 1561-76.
44 Berthon A, Drelon C, Ragazzon B et al. WNT/beta-catenin signalling is activated in aldosterone producing adenomas and controls aldosterone production. Hum Mol Genet 2014; 23: 889-905.
45 Tissier F, Cavard C, Groussin L et al. Mutations of beta-catenin in adrenocortical tumors: activation of the Wnt signaling pathway is a frequent event in both benign and malignant adrenocortical tumors. Cancer Res 2005; 65: 7622-7.
46 Assie G, Letouze E, Fassnacht M et al. Integrated genomic characterization of adrenocortical carcinoma. Nat Genet 2014; 46: 607-12.
47 Simpson ER, Waterman MR. Regulation of the synthesis of steroidogenic enzymes in adrenal cortical cells by ACTH. Annu Rev Physiol 1988; 50: 427-40.
48 Ross NS. Epidemiology of Cushing’s syndrome and subclinical disease. Endocrinol Metab Clin North Am 1994; 23: 539-46.
49 Reincke M. Subclinical Cushing’s syndrome. Endocrinol Metab Clin North Am 2000; 29: 43-56.
50 Beuschlein F, Reincke M, Karl M et al. Clonal composition of human adrenocortical neoplasms. Cancer Res 1994; 54: 4927-32.
51 Sidhu S, Marsh DJ, Theodosopoulos G et al. Comparative genomic hybridization analysis of adrenocortical tumors. J Clin Endocrinol Metab 2002; 87: 3467-74.
52 Kirschner LS, Sandrini F, Monbo J, Lin JP, Carney JA, Stratakis CA. Genetic heterogeneity and spectrum of muta- tions of the PRKAR1A gene in patients with the carney complex. Hum Mol Genet 2000; 9: 3037-46.
53 Kirschner LS, Carney JA, Pack SD et al. Mutations of the gene encoding the protein kinase A type I-alpha regulatory subunit in patients with the Carney complex. Nat Genet 2000; 26: 89- 92.
54 Zilbermint M, Stratakis CA. Protein kinase A defects and cortisol-producing adrenal tumors. Curr Opin Endocrinol Diabetes Obes 2015; 22: 157-62.
55 Beuschlein F, Fassnacht M, Assie G et al. Constitutive activation of PKA catalytic subunit in adrenal Cushing’s syndrome. N Engl J Med 2014; 370: 1019-28.
56 Cao Y, He M, Gao Z et al. Activating hotspot L205R mutation in PRKACA and adrenal Cushing’s syndrome. Science 2014; 344: 913-7.
57 Goh G, Scholl UI, Healy JM et al. Recurrent activating mutation in PRKACA in cortisol-producing adrenal tumors. Nat Genet 2014; 46: 613-7.
58 Sato Y, Maekawa S, Ishii R et al. Recurrent somatic muta- tions underlie corticotropin-independent Cushing’s syn- drome. Science 2014; 344: 917-20.
59 Di Dalmazi G, Kisker C, Calebiro D et al. Novel somatic mutations in the catalytic subunit of the protein kinase A as a cause of adrenal Cushing’s syndrome: a European multicen- tric study. J Clin Endocrinol Metab 2014; 99: E2093- 100.
60 Fassnacht M, Kroiss M, Allolio B. Update in adrenocortical carcinoma. J Clin Endocrinol Metab 2013; 98: 4551-64.
61 Abiven G, Coste J, Groussin L et al. Clinical and biological features in the prognosis of adrenocortical cancer: poor outcome of cortisol-secreting tumors in a series of 202 consecutive patients. J Clin Endocrinol Metab 2006; 91: 2650-5.
62 Allolio B, Fassnacht M. Clinical review: adrenocortical carci- noma: clinical update. J Clin Endocrinol Metab 2006; 91: 2027-37.
63 Libe R, Fratticci A, Bertherat J. Adrenocortical cancer: pathophysiology and clinical management. Endocr Relat Cancer 2007; 14: 13-28.
64 Arlt W, Biehl M, Taylor AE et al. Urine steroid metabolomics as a biomarker tool for detecting malignancy in adrenal tumors. J Clin Endocrinol Metab 2011; 96: 3775-84.
65 Fassnacht M, Terzolo M, Allolio B et al. Combination chemotherapy in advanced adrenocortical carcinoma. N Engl J Med 2012; 366: 2189-97.
66 Fraumeni JF Jr, Miller RW. Adrenocortical neoplasms with hemihypertrophy, brain tumors, and other disorders. J Pediatr 1967; 70: 129-38.
67 Custodio G, Parise GA, Kiesel Filho N et al. Impact of neonatal screening and surveillance for the TP53 R337H mutation on early detection of childhood adrenocortical tumors. J Clin Oncol 2013; 31: 2619-26.
68 Weksberg R, Smith AC, Squire J, Sadowski P. Beckwith- Wiedemann syndrome demonstrates a role for epigenetic control of normal development. Hum Mol Genet 2003; 12 Spec No 1: R61-8.
69 Wilkin F, Gagne N, Paquette J, Oligny LL, Deal C. Pediatric adrenocortical tumors: molecular events leading to insulin- like growth factor II gene overexpression. J Clin Endocrinol Metab 2000; 85: 2048-56.
70 West AN, Neale GA, Pounds S et al. Gene expression profiling of childhood adrenocortical tumors. Cancer Res 2007; 67: 600-8.
71 Gicquel C, Raffin-Sanson ML, Gaston V et al. Structural and functional abnormalities at 11p15 are associated with the malignant phenotype in sporadic adrenocortical tumors: study on a series of 82 tumors. J Clin Endocrinol Metab 1997; 82: 2559-65.
72 Giordano TJ, Thomas DG, Kuick R et al. Distinct transcrip- tional profiles of adrenocortical tumors uncovered by DNA microarray analysis. Am J Pathol 2003; 162: 521-31.
73 Drelon C, Berthon A, Ragazzon B et al. Analysis of the role of Igf2 in adrenal tumour development in transgenic mouse models. PLoS One 2012; 7: e44171.
74 Heaton JH, Wood MA, Kim AC et al. Progression to adreno- cortical tumorigenesis in mice and humans through insulin- like growth factor 2 and beta-catenin. Am J Pathol 2012; 181: 1017-33.
75 Fassnacht M, Berruti A, Baudin E et al. Linsitinib (OSI-906) versus placebo for patients with locally advanced or meta- static adrenocortical carcinoma: a double-blind, randomised, phase 3 study. Lancet Oncol 2015; 16: 426-35.
76 Herrmann LJ, Heinze B, Fassnacht M et al. TP53 germline mutations in adult patients with adrenocortical carcinoma. J Clin Endocrinol Metab 2012; 97: E476-85.
77 Else T. Association of adrenocortical carcinoma with familial cancer susceptibility syndromes. Mol Cell Endocrinol 2012; 351: 66-70.
78 Raymond VM, Everett JN, Furtado LV et al. Adrenocortical carcinoma is a lynch syndrome-associated cancer. J Clin Oncol 2013; 31: 3012-8.
79 Barreau O, de Reynies A, Wilmot-Roussel H et al. Clinical and pathophysiological implications of chromosomal alterations in adrenocortical tumors: an integrated genomic approach. J Clin Endocrinol Metab 2012; 97: E301-11.
80 Ragazzon B, Libe R, Gaujoux S et al. Transcriptome analysis reveals that p53 and {beta)-catenin alterations occur in a group of aggressive adrenocortical cancers. Cancer Res 2010; 70: 8276-81.
81 Barreau O, Assie G, Wilmot-Roussel H et al. Identification of a CpG island methylator phenotype in adrenocortical carcino- mas. J Clin Endocrinol Metab 2013; 98: E174-84.
82 de Reynies A, Assie G, Rickman DS et al. Gene expression profiling reveals a new classification of adrenocortical tumors and identifies molecular predictors of malignancy and sur- vival. J Clin Oncol 2009; 27: 1108-15.
83 Assie G, Jouinot A, Bertherat J. The ‘omics’ of adrenocortical tumours for personalized medicine. Nat Rev Endocrinol 2014; 10: 215-28.
84 Beuschlein F, Weigel J, Saeger W et al. Major prognostic role of Ki67 in localized adrenocortical carcinoma after complete resection. J Clin Endocrinol Metab 2015; 100: 841-9.
85 Pinto EM, Chen X, Easton J et al. Genomic landscape of paediatric adrenocortical tumours. Nat Commun 2015; 6:6302.
Correspondence: Per Hellman, Department of Surgical Sciences, University Hospital, 751 85 Uppsala, Sweden.
(fax: +4618-556808; e-mail: per.hellman@surgsci.uu.se).