EXPERT REVIEWS
Gene-expression profiling of adrenocortical carcinoma
Expert Rev. Mol. Diagn. 9(4), 343-351 (2009)
Insoo Suh, Marlon A Guerrero and Electron Kebebew+ *Author for correspondence University of California, San Francisco, Department of Surgery, Box 1674, San Francisco, CA 94143, USA Tel .: +1 415 885 3617 Fax: +1 415 885 7617 electron.kebebew@ ucsfmedctr.org
Adrenocortical carcinoma (ACC) is a rare malignancy of the adrenal cortex, associated with a generally dismal prognosis owing to its aggressive behavior. The clinical management of ACC is complicated by the inadequacy of current diagnostic modalities to accurately distinguish benign from malignant adrenocortical tumors. In addition, efforts to better predict clinical tumor behavior are limited by the lack of a better understanding of the molecular mechanisms of adrenocortical carcinogenesis. There have been no significant improvements in the treatment of ACC. Thus, there is a pressing need for the development of new therapeutic approaches for patients with ACC, as most patients present with advanced locoregional and metastatic disease. The prospects of identifying diagnostic and prognostic markers or gene profiles for ACC have significantly improved with the development of genome-wide gene-expression analysis. Since 2003, several studies have reported distinct gene-expression profiles between benign and malignant adrenocortical tumors that may have diagnostic and prognostic clinical utility. In this article, we discuss the limitations of the clinical management of ACC, and the gene-expression profile studies that have attempted to address these limitations.
KEYWORDS: adrenal incidentaloma · adrenalectomy · adrenocortical carcinoma · diagnostic marker · gene-expression profiling · prognostic marker
The ability to analyze expression levels of thou- sands of genes simultaneously using genome- wide cDNA microarrays has revolutionized the ways in which investigators could approach ques- tions on the molecular genetics of cancer. The most common application of this technology in cancer research is the comparison of micro- array data, or ‘profiles’, between malignant and benign tumors of the same tissue type, which generates a group of genes that have significant differences in expression level between the two groups. Under the assumption that these genes have some biologic relevance to the mechanisms of carcinogenesis, investigators could use micro- array data as ‘hypothesis builders’ when design- ing further mechanistic or functional genomic experiments in cancer models. Alternatively, gene-expression profiling of a variety of human malignancies also has potential clinical appli- cation as diagnostic and predictive markers of disease and in defining molecular classes of tumors that have traditionally been classified by tumor architecture and morphology or cytol- ogy. Moreover, gene-expression profile studies have also helped identify potential targets for therapies when confirmed to be biologically relevant in tumor cell biology. Genome-wide
gene-expression profiling has been used to analyze a variety of human cancers, including adrenocortical carcinoma (ACC). It allows for the meaningful correlation of molecular profiles with clinical variables, the classification or defi- nition of different tumor types and the identi- fication of genes or networks of genes involved in carcinogenesis.
Adrenocortical carcinoma is a rare but aggres- sive cancer of the adrenal cortex, with a dismal prognosis. The estimated annual incidence is approximately one case per million people [1-5]. The poor outcome in most patients with ACC is due, in part, to late stages of disease at diag- nosis, since approximately two-thirds of patients present with advanced locoregional disease and distant metastasis [6]. Even patients with local- ized ACC are difficult to cure because over half of patients will develop recurrent or persistent disease after apparent complete surgical resec- tion [2,3,6,7]. More-effective therapeutic strategies for ACC are desperately needed because of the high mortality and aggressive behavior associated with this disease.
The management of an incidentally discov- ered adrenal tumor (adrenal incidentaloma) is a common finding on cross-sectional radiographic
imaging studies, and represents a growing clinical dilemma with- out a definitive optimal management strategy established [8]. There is no reliable clinical evaluation to exclude ACC in the growing number of patients with localized adrenal incidentaloma. While gene-expression profiling studies of ACCs is relatively young by comparison to other more common malignancies, its findings thus far already have exciting potential for diagnostic and prognostic use in the clinical management of ACC. In this article, we provide an overview of the clinical limitations in the management of ACC, the currently known knowledge of adreno- cortical carcinogenesis and gene-expression profiling studies that have attempted to improve upon these limitations.
Clinical management of ACC Diagnosis
The dismal outlook of ACC is due, in part, to the late stage of presentation in most patients. At the time of diagnosis, approxi- mately 57% of patients present with local extension or distant metastasis [9]. The reason for this is that there are no specific symptoms diagnostic of ACC, and most patients present with vague abdominal complaints that cannot clearly be ascribed to adrenal origin. These features contribute to the challenge of diag- nosing ACC at an earlier stage. Although adrenal incidentalomas are being detected more frequently because of the wide use of sensitive imaging studies, the extent of disease at diagnosis for ACC has not changed in the last 40 years [9].
There are three general patterns of clinical presentation for ACC: an adrenal incidentaloma, a biochemically active adrenal mass causing a hormonal hypersecretion syndrome, or an adrenal mass
causing vague abdominal symptoms or mass effect. The prevalence of adrenal incidentaloma is approximately 5% and is higher in the elderly population [9]. Approximately 60% of patients with ACC present with evidence of steroid hormone hypersecretion [7]. The most common presentation is Cushing’s syndrome, occurring in up to a third of cases [5,7,9]. While most hormonal syndromes may be distributed equally among benign and malignant adrenal tumors, virilizing tumors in women and feminizing tumors in men are more commonly associated with ACC. It is, therefore, important to assess all adrenal tumors for hormonal hypersecretion.
Evaluation of any adrenal tumor, including an adrenal inciden- taloma, should proceed in a systematic manner to assess for bio- chemical activity. Screening for Cushing’s syndrome involves an overnight dexamethasone (1-mg) suppression test. A cortisol level below 5 µg/dl excludes Cushing’s syndrome. Elevated cortisol necessitates checking a 24-h urinary cortisol, plasma adrenocor- ticotropic hormone level and a high-dose dexamethasone (8-mg) suppression test [7,10]. Virilizing or feminizing tumors are diag- nosed by serum measurement of sex steroids and their precursors: testosterone, dehydroepiandrosterone sulfate, androstenedione, 17-hydroxy-progesterone and 17ß-estradiol. An elevated level of dehydroepiandrosterone sulfate is more commonly associated with ACC; however, even ACCs that are not symptomatic of hormonal hypersecretion are associated with higher levels of androstenedi- one or 17-hydroxy-progesterone [7]. A 24-h urine collection and measurement of fractionated catecholamines and metanephrines is always obtained to rule out a pheochromocytoma. In particu- lar, measurement of fractionated plasma-free metanephrines and normetanephrine is highly accurate for diagnosing a pheochro- mocytoma, with a sensitivity of 99% and specificity of 89% [8]. Primary hyperaldosteronism should also be excluded in patients with concurrent hypertension and or hypokalemia. A normo- tensive patient with normal serum potassium generally excludes primary hyperaldosteronism. In patients with hypertension, a plasma aldosterone and renin activity is determined, and the ratio of aldosterone-to-renin is calculated. A ratio of over 30 and a plasma aldosterone level of over 20 ng/dl confirms the presence of an aldosteronoma [8]. Aldosteronomas are rarely malignant but may present with excess secretion of sex hormones or cortisol.
In addition to the biochemical evaluation, patients with adrenal masses undergo adrenal-protocol cross-sectional imaging studies (2-3-mm sections from the lower chest to the bifurcation of the aorta) to evaluate for tumor size and other features suggestive of malignancy, such as irregular borders, heterogeneity, stippled calcifications, necrosis, local tissue invasion, regional lymph- adenopathy or distant metastasis (FIGURE 1). The preferred imag- ing modality is a high-resolution CT scan of the abdomen. The lipid content of an adrenal mass, as measured by determining the Hounsfield units (HU), as well as the timing of contrast wash- out, can help distinguish between benign and malignant tumors. Typically, adrenal adenomas have a density of or below 10 HU on unenhanced CT and below 30 HU on contrast-enhanced CT with rapid washout of contrast equal to or greater than 50% at 10 min. ACCs are typically vascular, with slow washout (<50%) and have a HU of great than 10 on unenhanced computed tomography [11].
MRI of the abdomen, although more expensive than a CT scan, is also utilized to evaluate adrenal tumors. Features on MRI that suggest malignancy include heterogeneous signal intensity on T1- and T2-weighted imaging, peripheral nodular enhancement and central hypoperfusion on contrast MRI [5].
The tumor size of adrenal neoplasms measured by imaging stud- ies has been used as a preoperative surrogate marker for malig- nancy and for recommending surgical resection. However, the appropriate management of adrenal tumors that measure 4-6 cm is controversial [8,12-14]. Even when the tumor is 6 cm or larger, the reported risk of malignancy ranges from 5 to 98%, depend- ing on the study cohort [15]. In many centers, a size threshold of greater than 6 cm has been used as an indication for adrenalec- tomy. However, it is unclear whether adrenal tumors between 4 and 6 cm should be removed or monitored, whereas most experts recommend monitoring tumors less than 4 cm in size. Apart from whether adrenalectomy is indicated, whether an open or laparo- scopic approach should be used has depended on the likelihood of malignancy, technical issues and the experience of the surgeon. In a recent double-cohort study comparing tumor size of local- ized benign and malignant adrenocortical tumors, the sensitivity, specificity and likelihood ratio of tumor size in predicting malig- nancy were 96%, 52% and 2.0, respectively, for tumors equal to or greater than 4 cm; 90%, 80% and 4.4 for tumors equal to or greater than 6 cm; 77%, 95% and 16.9 for tumors equal to or greater than 8 cm; and 55%, 98% and 24.4 for tumors equal to or greater than 10 cm (FIGURE 2) [15]. Although these results suggest that tumor size is a useful surrogate marker for ACCs and may help surgeons choose an open resection, it would still subject many patients to an unnecessary operation, as well as miss small ACCs.
In 2002, the NIH held a State of the Science Conference to estab- lish guidelines for the management of adrenal incidentalomas. The resulting recommendations suggested that all nonfunction- ing incidentalomas of over 6 cm be resected, and that masses below 4 cm be observed with repeat imaging to monitor for growth at 6-month follow-up [8]. No recommendation for resec- tion versus observation was made for adrenal tumors between 4 and 6 cm because of a lack of strong clinical evidence. However, a recent study found that tumors over 4 cm are still associated with a doubled risk of malignancy. Furthermore, the tumor size measurements based on imaging studies may underestimate size, especially in cases of smaller tumors. Owing to these factors, it is currently recommended that nonfunctioning tumors larger than 4 cm in size should be resected [15].
Treatment
Therapeutic options for ACC are usually determined according to the stage at presentation based on imaging studies, and include surgical resection, mitotane, adjuvant chemotherapy, and/or external beam radiation (TABLE 1). The standard treatment recom- mendation for all patients with localized ACC (stage I and II) is open adrenalectomy if there is clear evidence of malignancy and laparoscopic adrenalectomy if the tumor is small and there is no clear evidence of malignancy that can be established pre- operatively or at the time of exploration. Stage III disease is also
Tumor size distribution (%)
70
Malignant
60
Benign
50
40
30
20
10
0
<2
2-3.9
4-5.9
6-7.9
8-9.9
>10
Tumor size (cm)
approached surgically; however, a regional lymph node dissection is recommended for those who present with evidence of lymph node involvement. Due to the high risk of recurrence and poor prognosis, consideration is often given to enrolling patients in a clinical trial for adjuvant therapy. Stage IV disease is approached nonoperatively and is primarily treated with mitotane with or without chemotherapy. Radiation is employed in cases of bone metastasis. Surgery may be considered in cases of localized metas- tasis with no evidence of widely disseminated disease. Clinical trials are also considered in patients with stage IV ACC.
Mitotane has been utilized in the treatment of ACC since the US FDA approved its use in 1970 [16]. Mitotane has a specific cyto- toxic effect on the adrenal gland that causes focal degeneration predominantly in the fascicular and reticular zones of the adrenal cortex. It has been shown that mitotane leads to tumor regression in 25% of cases with a targeted plasma-concentration window of 14-20 mg/l [5,7]. The difficulty with its use is its narrow therapeutic window and the frequency of adverse effects involving the GI tract and CNS. Furthermore, adrenal suppression from mitotane necessi- tates glucocorticoid supplementation [7]. Overall, although mitotane is recommended in patients with advanced or metastatic disease, its low and variable efficacy and frequent side effects have precluded a consensus on its adjuvant use. In addition, there is controversy regarding the actual benefit of mitotane on disease-free survival and overall survival. As such, adjuvant use of mitotane in stage I and II ACCs was not recommended during a recent international consensus conference for the management of ACC [5].
Recently, a multi-institutional study in Italy and Germany com- pared patients treated with radical surgery and adjuvant mitotane to patients treated with surgery alone. The investigators found that disease-free survival was prolonged in the mitotane group, compared with two control groups of surgery alone (42 vs 10 and 25 months; p < 0.05). After adjusting for age, sex and stage, the control groups had a higher risk of recurrence and death compared with the mitotane group [17]. The importance of this study is that this difference was achieved at low doses (1-5 g daily) with a lower frequency of high-grade adverse effects, thus opening the possibility for more widespread use.
Review
| Table 1. TNM staging of adrenocortical carcinoma. | |
|---|---|
| Stage | Tumor characteristics |
| I | <5 cm |
| II | >5 cm without invasion |
| III | Any size with local fat extension or lymph node involvement |
| IV | Invasion of adjacent organs or distant metastasis |
| TNM: Tumor, node, metastases. | |
There is no current cytotoxic chemotherapy protocol that has proven beneficial in long-term survival. This is, in part, due to the rarity of this malignancy and the resultant limitations in thoroughly evaluating many different chemotherapy regimens. Nonetheless, there have been studies that have shown promising responses to chemotherapy regimens. A protocol devised by Berruti et al. combines etoposide, cisplatin, doxorubicin and mitotane. This regimen had an overall response rate of close to 50% [18]. Khan et al. demonstrated an overall response rate of 36% with a combination of mitotane and streptozocin [19]. These results have prompted a Phase III clinical trial to compare the two regimens in the First International Randomized Trial in Locally Advanced and Metastatic Adrenocortical Carcinoma Treatment (FIRM- ACT) trial, which continues to actively accrue patients in Europe and USA. Recently, an interesting retrospective study compared the efficacy of various chemotherapy agents used to treat ACC and concluded that there was no difference between mitotane alone and combination chemotherapy [20]. This study had several limi- tations, so the pending results from the FIRM-ACT trial will be very important. Currently, combination chemotherapy protocols with mitotane are currently recommended for advanced stage III and IV disease, as well as recurrent disease. As in single therapy with mitotane, glucocorticoid supplementation should be utilized in the presence of adrenal insufficiency. Even though adjuvant or primary mitotane and chemotherapy may have some efficacy, the disease-free and overall survival remains poor for patients with ACC and new therapeutic strategies are desperately needed.
The efficacy of external-beam radiation as adjuvant therapy after complete resection has not been proven. Radiation therapy is cur- rently recommended for the treatment of metastasis to the bone and brain, as well as for symptomatic unresectable local recurrences and cases in which an R0 resection could not be successfully per- formed during the initial operation. In addition, radiation has been used after resection of isolated local recurrences, but the efficacy of such an approach is unclear [5]. Since no studies have clearly defined efficacy of radiation therapy in unresectable ACC, its use in patients with asymptomatic, unresectable local recurrences should be considered individually.
Limitations in ACC management
As described, there are no reliable preoperative clinical, imaging or biochemical tests to distinguish between primary benign and malignant adrenocortical neoplasms in the absence of obvious metastatic disease or locoregional invasion [14]. Imaging features,
such as large tumor size, heterogeneity, irregular border, hem- orrhage, necrosis, rapid tumor growth rate and a score greater than 10-20 HU on a noncontrast CT scan are more common in malignant tumors but are not reliable enough to use for making definitive treatment decisions [14,21]. In patients with a history of an extra-adrenal malignancy, an adrenal incidentaloma may indicate metastatic disease in 32-73% of cases [22,23]. Fine-needle aspira- tion biopsy is inaccurate in most cases for distinguishing primary benign from malignant adrenocortical tumors, although it may be useful for detecting metastatic disease to the adrenal gland after biochemical exclusion of a hyperfunctioning tumor [13,24-28].
Histopathological diagnosis of ACC is also difficult. Several features that have been proposed suggest malignancy, including macroscopic and microscopic features, but none have been uni- formly accepted. The most commonly used histologic criteria for ACC are those developed by Weiss [29]. This scoring model uses nine histologic features to determine the malignant potential of adrenocortical tumors: high mitotic count, high nuclear grade, atypical mitoses, necrosis, capsular invasion, vascular invasion, sinusoidal invasion, diffuse architecture and nonclear cytoplasm. With each criterion given a score of 1, a score of 3 or above has been reported to have a sensitivity of 100% and specificity of 96% [30]. However, the Weiss scoring system cannot definitively predict malignancy, especially considering that ACCs have been found in tumors with a score of less than 2. The Weiss score is further limited by the subjectivity of some of the criteria, which are prone to sampling and human error [31]. Tumor staining with Ki-67 has been utilized to help differentiate malignant from benign adrenal tumors, with limited success [7]. New diagnostic approaches for distinguishing benign from malignant adrenocortical tumor are needed to avoid unnecessary surgical resection and for select- ing the appropriate surgical approach. Additionally, markers that could predict which patients have aggressive disease and would benefit from adjuvant therapy are also needed.
Most patients with ACC have a dismal prognosis and there has been no significant improvement in the outcome of patients in over 50 years. The growing number of adrenal incidentalomas may have led to earlier diagnosis of ACC and, thus, better outcome. However, no significant difference in the incidence, tumor size or extent of disease at diagnosis and treatment has been observed for 40 years. There is a pressing need for the development of new therapeutic approaches for patients with ACC, as most patients present with advanced locoregional and metastatic disease.
Genetics of ACC
The molecular basis of adrenocortical carcinogenesis is poorly understood. Most studies have focused on the genetic changes associated with hereditary cancer syndromes, of which ACC is a component. These include Beckwith-Wiedemann syndrome (associated with germline 11p15 chromosomal alterations leading to IGF2 overexpression), Li-Fraumeni syndrome (TP53 muta- tion), multiple endocrine neoplasia type 1 (mutations in the menin tumor-suppressor gene), and Gardner’s syndrome (APC mutation) [31,32]. Although the majority of ACCs are sporadic and not hereditary tumors, investigations of the genetic changes
associated with hereditary cancer syndromes and genome-wide chromosomal aberration studies have provided important infor- mation about the carcinogenesis in a subset of sporadic ACCs. In fact, all of the genetic alterations mentioned have been found to be more frequent in sporadic ACCs compared with benign adrenocortical adenomas, with some alterations such as loss of heterozygosity (LOH) of 11q13 occurring in up to 100% of ACCs [33-36].
Genome-wide assessment of ACCs first began in the form of LOH analysis and comparative genomic hybridization (CGH) studies. These powerful techniques were able to identify larger chromosomal region changes in the form of copy-number differ- ences (in the case of CGH) or losses in certain alleles (in LOH). LOH studies were critical in identifying the key chromosomal alterations in the hereditary tumor syndromes described earlier, as well as several studies on sporadic ACCs [32]. Early CGH studies definitively established clear genetic differences in ACCs by show- ing that they harbor up to 14-fold more copy-number changes compared with benign tumors, with larger tumors also having a greater number of changes [37,38]. In general, the most common chromosomal gains in ACC were found in chromosomes 4, 5, 12 and 19, while losses were most commonly found in 1, 2, 3, 4, 6, 9, 11, 13, 15, 17, 18, 22 and X [37]. While valuable as a general survey, neither the CGH nor LOH analyses are able to identify specific genes or mechanisms responsible for cancer formation and progression; furthermore, their clinical application as diagnostic or prognostic tools has been limited for ACC.
Gene-expression profiling studies in ACC
The identification of specific diagnostic and prognostic molecu- lar markers of malignancy, as well as a greater elucidation of the mechanisms of adrenocortical carcinogenesis, remains a pressing issue if the diagnostic accuracy and outcome of patients with ACC is to be improved. In an attempt to address both of these goals, several groups (including our own) have applied high-throughput gene-expression microarray analysis to the study of ACCs (TABLE 2). Giordano et al. published the first microarray study of sporadic ACCs in 2003. This group from the University of Michigan (USA), used an Affymetrix (CA, USA) HG_U95Av2 array cover- ing approximately 10,500 genes to compare the expression profiles of 11 ACCs to those of eight benign adrenocortical tissue samples. Using a cutoff of above threefold differential expression with a F-score p-value of less than 0.01, they identified 91 differentially expressed (DE) genes, 41 of which were overexpressed in ACC and 50 of which were underexpressed. Notably, the most over- expressed gene overall was IGF2, which was already known to be relevant from its role in Beckwith-Wiedemann syndrome; it was found on this microarray analysis to be overexpressed more than 100-fold in ACC compared with benign adrenocortical tissues. This finding was further validated using quantitative PCR on the same tissue samples [39].
A total of five additional groups, including our own, have also studied ACC by microarray analysis, with varying microarray technologies, statistical methodologies and resultant differentially expressed genes identified (TABLE 2) [40-45]. All of the studies from
the USA used arrays produced by Affymetrix, but a number of groups from Europe created custom arrays from their own institu- tions or from cooperating consortiums; furthermore, the specific techniques used for hybridization and normalization of raw data also differed depending on the institutions’ specific protocol. The statistical thresholds for determining significance in differential expression were also of varying stringency. Of all of the mentioned groups, only West et al. specifically crossvalidated their microarray platform with those of other groups, in an attempt to find concor- dance in gene-expression results across different arrays [44]. These issues of variability in design and execution are present in any microarray analysis, and are the subject of vigorous discussion in the genomics and bioinformatics communities, with the eventual goal of establishing protocols for standardizing microarray data acquisition and analysis [46,47].
Despite the differences in the result of the microarray studies, the IGF2 gene was specifically noted in nearly all of the stud- ies to be overexpressed in ACCs, and deserves special mention here as an important mediator in adrenocortical carcinogenesis. The IGF family consists of a complex network of regulatory proteins that mediate cell proliferation and survival [48]. Both IGF1 and IGF2 are activating ligands for IGF1R, a cell-surface tyrosine kinase receptor that serves to promote the downstream signaling of both the PI3K/Akt and RAS/MAPK intracellular pathways. Interestingly, IGF2 also binds another receptor in the family, IGF2R, but this does not appear to lead to any downstream signaling activation and is currently thought to be a competitive inhibitor of IGF1R-mediated proliferation by reducing the amount of IGF2 available. The overexpression of IGF2, therefore, could presumably overwhelm the nega- tive regulatory effects of IGF2R and lead to an uncontrolled activation of the IGF1R-mediated pathways, thus leading to malignant transformation.
As mentioned, IGF2 gene alterations as a result of LOH of chromosome 11p15 had already been characterized in heredi- tary Beckwith-Wiedemann syndrome. Interestingly, sporadic ACC cases frequently harbor these alterations as well but as a somatic change. LOH of 11p15 has been shown to occur in up to 83% of sporadic ACCs versus 34% of adrenocortical adenomas, and overexpression of IGF2 has been found in up to 60-90% of ACCs [49,50]. Among the microarray findings, only one study by Lombardi et al. did not specifically make any mention of IGF2, possibly because it was not one of the 82 measured genes on their more-limited array platform [43]. All of the remaining five genome-wide studies specifically mentioned their findings on IGF2, and all showed at least some degree of overexpression in ACCs [39-42,44,45]. The apparent relevance of IGF2 in adre- nocortical carcinogenesis has important therapeutic potential, specifically by tailoring or developing therapies that are specifi- cally designed to inhibit IGF2-mediated mitogen signaling. In fact, a new Phase II trial for patients with ACC has opened to evaluate the efficacy of an antibody (recombinant human IgG1 monoclonal, IMC-A12) targeting IGF-1R because of promising preclinical studies that showed a significant antineoplastic effect from inhibiting IGF signaling [51,52].
| Study (year) | Comparison groups | Microarray platform | Probe sets (n) | Criteria for determining differential expression | Number of differentially expressed genes identified | IGF2 overexpression? | Ref. |
|---|---|---|---|---|---|---|---|
| Giordano et al. (2003) | 11 ACCs 8 benign specimens (4 ACAS, 3 normal adrenal cortices, 1 macronodular hyperplasia) | GeneChip® HG_U95Av2 (Affymetrix, CA, USA) | 12,625 | >3-fold expression difference and F-test p < 0.01 | 91 (41 overexpressed, 50 underexpressed) | Yes | [39] |
| de Fraipont et al. (2005) | 24 ACCs 33 ACAS | Custom platform from the IMAGE consortium | 230 | t-test (p-value not specified) | 22 | Yes | [40] |
| Velázquez- Fernández et al. (2005) | 7 ACCs 13 ACAS | KTH HUM 30 k cDNA (KTH Royal Institute of Technology, Stockholm, Sweden) | 29,760 | >2-fold expression difference and B-test B > 0 and t-test p < 0.01 | 571 (273 overexpressed, 298 underexpressed) | Yes | [41] |
| Slater et al. (2006) | 10 ACCs 10 ACAS | IMT 12 k human cDNA (Philipps University, Marburg, Germany) | 11,540 | >2-fold expression difference and t-statistic absolute | 74 (19 overexpressed, 55 underexpressed) | Yes | [42] |
| Lombardi et al. (2006) | 2 ACCs 2 ACAS | Atlas® cDNA expression (BD Biosciences Clontech, CA, USA) | 82 | >1.5-fold expression difference | 4 (3 overexpressed, 1 underexpressed) | Not specified | [43] |
| West et al. (2007) | 18 ACCs 5 ACAS | GeneChip U133A (Affymetrix) | 22,215 | Wilcoxon rank-sum test p < 0.001 | 52 (23 overexpressed, 29 underexpressed) | Yes | [44] |
| Fernandez- Ranvier et al. (2008) | 5 ACCs 74 ACAs | GeneChip U133 Plus 2.0 (Affymetrix) | >47,000 | >8-fold expression difference and false discovery rate < 5% and adjusted t-test p < 0.01 | 37 (22 overexpressed, 15 underexpressed) | Yes | [45] |
| de Reynies et al. (2009) | 92 ACCs and ACAs | GeneChip U133 Plus 2.0 | >47,000 | Unsupervised cluster analysis | 746 probe sets | Yes | [55] |
| Giordano et al. (2009) | 33 ACCs 22 ACAS 10 normal adrenal cortices | GeneChip U133 Plus 2.0 | >47,000 | >1.5-fold expression difference p < 0.001 | 1890 (879 overexpressed, 1011 underexpressed) | Yes | [56] |
| ACA: Adrenocortical adenoma; ACC: Adrenocortical carcinoma; IMAGE: Integrated Molecular Analysis of Gene Expression. | |||||||
Clinical applications of microarray analysis of ACC
It is clear that IGF2 plays an important role in the mechanism of adrenocortical carcinogenesis. Whether it plays a universal role - and by extension, can be used as an accurate diagnostic marker - is less clear. Indeed, the diagnostic potential of IGF2 expression for predicting ACC has been somewhat conflicting, with varying sensitivities of 60-77% [31,53]. Furthermore, none of the earlier ACC microarray studies specifically addressed the issue of whether the expression of any of the DE genes that were identified could be applied for diagnostic use.
Our group attempted to address this question in our micro- array studies. In a genome-wide comparison of five ACCs ver- sus 74 benign adrenocortical adenomas, we identified 37 DE genes using relatively stringent criteria for differential expression. Notably, we found IGF2 to be overexpressed in ACCs but not high enough to meet our statistical criteria, which may more accurately reflect the relevant but varied role that IGF2 actually plays in ACC. After validating the expression of these genes using quantitative PCR, we tested the diagnostic accuracy of the most significant DE genes in predicting malignancy. We found that a
combination five seemingly unrelated genes, IL13RA2, HTR2B, CCNB2, RARRES2 and SLC16A9, had an overall accuracy of 91% [45].
In a separate study, we also performed a subanalysis of our microarray data to identify only those DE genes that are located on chromosome 11q13, a region with a high frequency of LOH in ACC. As an interesting possible confirmation of the existing knowledge of the relationship between ACC and 11q13 LOH, we found that all 21 of the discovered DE genes on microarray analysis were underexpressed in ACCs. After validating their expression with quantitative PCR, the six most underexpressed genes (SERPING1, MRPL48, TM7SF2, DDB1, NDUFS8 and PRDX5) had an overall accuracy of 89% when used in combina- tion [54]. To our knowledge, our studies are the first to perform formal analyses of diagnostic accuracy for ACC by validating and formally analyzing the accuracy for distinguishing benign from malignant adrenocortical neoplasms.
In addition to its use as diagnostic adjuncts, microarray profiles have also been evaluated as a predictor of prognosis for disease states. In ACC, several groups have evaluated this question. De Fraipont et al. from France performed a subanalysis of their microarray data, focusing only on the 13 ACCs that were not metastatic at the time of surgical resection. Despite the relatively small number of tumors/patients, supervised clustering analysis nevertheless identi- fied a group of 14 genes that were able to discriminate between recurrent and nonrecurrent cancers. The expression levels of the two most DE genes, ISGF3G and Fos, were shown on Kaplan-Meier survival analysis to predict disease-free survival with 100% accu- racy at 60 months [40]. A recent study by Reynies and colleagues of 153 adrenocortical tumors identified three distinct groups of tumor on unsupervised analysis based on the presence of metastasis, relapse and cause-specific mortality [55]. They identified that a combined two-gene signature (DLG7 and INK1) predicted disease-free survival and overall survival. They validated their results in an independent cohort by quantitative real-time-PCR. Also, in a follow-up study, Giordano et al. studied ten normal adrenal cortex, 22 adrenocortical adenoma and 33 ACC using the Affymetrix Human Genome U133 Plus 2.0 oligonucletoide array [56]. In this study using principal com- ponent analysis, they found good separation between high-grade and low-grade ACCs. Furthermore, cluster analysis showed two clusters of tumor related to proliferation as determined by mitotic counts and cell cycle genes. On multivariate analysis including stage (I and II vs III and Key issues IV), mitotic rate and gene expression data were independently predictive of prognosis. Results such as these could foreseeably lead to treatment regimens that are tailored to a specific ACC molecular marker profile, and further highlight the promising therapeutic potential of microarray applications.
Expert commentary
Adrenocortical carcinoma is a challeng- ing cancer to diagnose and treat, and new modalities are necessary if the management
of affected patients is to improve. Although histology is the clini- cal standard for diagnosing localized ACC, the clinical utility for making management decisions is limited. Thus, new classifica- tion systems besides histology need to be considered. Genome- wide expression studies in ACC suggest that molecular classi- fication may be possible but a uniform panel of diagnostic and prognostic markers has not yet been established. Combining the genome-wide gene-expression studies would not be appropriate because of the different methodologies used. However, a pathway analysis-based approach may identify the best panel of mark- ers to distinguish benign from malignant adrenocortical tumors and predict ACC behavior. We believe that future studies in larger cohorts and high-quality clinical data will be needed to be undertaken to arrive at clinically robust diagnostic and prog- nostic markers for ACC. The standardization techniques and analytic methods for genome-wide gene-expression studies will also greatly enhance the likelihood of arriving at clinical use- ful markers for ACC. These goals are most likely to be realized through the establishment of cooperative study groups because ACC is such a rare malignancy.
Five-year view
All of the gene-expression profile studies have been based on retrospectively collected adrenocortical tumors, with only few studies that have been crossvalidated or validated by quantita- tive PCR. In the future, it is expected that prospective validation of the gene-expression profiles in clinical samples could lead to the identification of reliable diagnostic and prognostic markers for ACC. Moreover, an integrated approach of using microarray data with high-throughput epigenetic, microRNA, proteomic and comparative genomic hybridization analysis in ACC could result in a better understanding of adrenocortical carcinogenesis and the development of diagnostic and prognostic markers for ACC.
Financial & competing interests disclosure
The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.
No writing assistance was utilized in the production of this manuscript.
· The diagnosis of localized adrenocortical carcinoma (ACC) is imperfect, even when based on histology.
· ACC is associated with a poor prognosis and there is no reliable criteria for selecting which subset of patients will benefit from primary or adjuvant mitotane and chemotherapy.
· Multiple diagnostic molecular signatures for ACC have been reported in resected ACC.
· Except for IGF expression, there is a discrepancy in the specific genes that distinguish benign from malignant adrenocortical tumors.
· Very few studies have validated the diagnostic gene-expression signatures reported.
· The performance of these gene-expression profiles to clinical diagnostic criteria for ACC is undetermined.
· There are limited data on gene-expression profiles that predict ACC prognosis.
Review
References
Papers of special note have been highlighted as:
· of interest
·· of considerable interest
1 Gonzalez RJ, Tamm EP, Ng C et al. Response to mitotane predicts outcome in patients with recurrent adrenal cortical carcinoma. Surgery 142(6), 867-875; discussion 867-875 (2007).
2 Icard P, Goudet P, Charpenay C et al. Adrenocortical carcinomas: surgical trends and results of a 253-patient series from the French Association of Endocrine Surgeons study group. World J. Surg. 25(7), 891-897 (2001).
3 Kendrick ML, Lloyd R, Erickson L et al. Adrenocortical carcinoma: surgical progress or status quo? Arch. Surg. 136(5), 543-549 (2001).
4 Paton BL, Novitsky YW, Zerey M et al. Outcomes of adrenal cortical carcinoma in the United States. Surgery 140(6), 914-920; discussion 919-920 (2006).
5 Schteingart DE, Doherty GM, Gauger PG et al. Management of patients with adrenal cancer: recommendations of an international consensus conference. Endocr. Relat. Cancer 12(3), 667-680 (2005).
6 Sidhu S, Sywak M, Robinson B, Delbridge L. Adrenocortical cancer: recent clinical and molecular advances. Curr. Opin. Oncol. 16(1), 13-18 (2004).
7 Allolio B, Fassnacht M. Clinical review: adrenocortical carcinoma: clinical update. J. Clin. Endocrinol. Metab. 91(6), 2027-2037 (2006).
8 NIH state-of-the-science statement on management of the clinically inapparent adrenal mass (“incidentaloma”). NIH Consens. State Sci. Statements 19(2), 1-25 (2002).
9 Kebebew E, Reiff E, Duh QY, Clark OH, McMillan A. Extent of disease at presentation and outcome for adrenocortical carcinoma: have we made progress? World J. Surg. 30(5), 872-878 (2006).
10 Young WF Jr. Clinical practice. The incidentally discovered adrenal mass. N. Engl. J. Med. 356(6), 601-610 (2007).
11 Caoili EM, Korobkin M, Francis IR et al. Adrenal masses: characterization with combined unenhanced and delayed enhanced CT. Radiology 222(3), 629-633 (2002).
12 Mansmann G, Lau J, Balk E, Rothberg M, Miyachi Y, Bornstein SR. The clinically inapparent adrenal mass: update in diagnosis and management. Endocr. Rev. 25(2), 309-340 (2004).
13 Shen WT, Sturgeon C, Duh QY. From incidentaloma to adrenocortical carcinoma: the surgical management of adrenal tumors. J. Surg. Oncol. 89(3), 186-192 (2005).
14 Sturgeon C, Kebebew E. Laparoscopic adrenalectomy for malignancy. Surg. Clin. North Am. 84(3), 755-774 (2004).
15 Sturgeon C, Shen WT, Clark OH, Duh QY, Kebebew E. Risk assessment in 457 adrenal cortical carcinomas: how much does tumor size predict the likelihood of malignancy? J. Am. Coll. Surg. 202(3), 423-430 (2006).
16 Lubitz JA, Freeman L, Okun R. Mitotane use in inoperable adrenal cortical carcinoma. JAMA 223(10), 1109-1112 (1973).
17 Terzolo M, Angeli A, Fassnacht M et al. Adjuvant mitotane treatment for adrenocortical carcinoma. N. Engl. J. Med. 356(23), 2372-2380 (2007).
18 Berruti A, Terzolo M, Sperone P et al. Etoposide, doxorubicin and cisplatin plus mitotane in the treatment of advanced adrenocortical carcinoma: a large prospective Phase II trial. Endocr. Relat. Cancer 12(3), 657-666 (2005).
19 Khan TS, Imam H, Juhlin C et al. Streptozocin and o,p’DDD in the treatment of adrenocortical cancer patients: long-term survival in its adjuvant use. Ann. Oncol. 11(10), 1281-1287 (2000).
20 Fareau GG, Lopez A, Stava C, Vassilopoulou-Sellin R. Systemic chemotherapy for adrenocortical carcinoma: comparative responses to conventional first-line therapies. Anticancer Drugs 19(6), 637-644 (2008).
21 Duh QY. Functioning and non- functioning adrenal tumors. In: The Practice of General Surgery. Bland KI (Ed.). WB Saunders, PA, USA, 1077-1082 (2002).
22 Kebebew E, Siperstein AE, Clark OH, Duh QY. Results of laparoscopic adrenalectomy for suspected and unsuspected malignant adrenal neoplasms. Arch. Surg. 137(8), 948-951; discussion 952-943 (2002).
23 Kebebew E, Duh QY. Adrenal incidentaloma. In: Current Surgical Therapy. Cameron JL (Ed.). Elsevier Mosby, PA USA, 575-579 (2004).
24 Heaston DK, Handel DB, Ashton PR, Korobkin M. Narrow gauge needle aspiration of solid adrenal masses. AJR Am. J. Roentgenol. 138(6), 1143-1148 (1982).
25 Berkman WA, Bernardino ME, Sewell CW, Price RB, Sones PJ Jr. The computed tomography-guided adrenal biopsy. An alternative to surgery in adrenal mass diagnosis. Cancer 53(10), 2098-2103 (1984).
26 Katz RL, Patel S, Mackay B, Zornoza J. Fine needle aspiration cytology of the adrenal gland. Acta Cytol. 28(3), 269-282 (1984).
27 Welch TJ, Sheedy PF 2nd, Stephens DH, Johnson CM, Swensen SJ. Percutaneous adrenal biopsy: review of a 10-year experience. Radiology 193(2), 341-344 (1994).
28 Silverman SG, Mueller PR, Pinkney LP, Koenker RM, Seltzer SE. Predictive value of image-guided adrenal biopsy: analysis of results of 101 biopsies. Radiology 187(3), 715-718 (1993).
29 Weiss LM, Medeiros LJ, Vickery AL Jr. Pathologic features of prognostic significance in adrenocortical carcinoma. Am. J. Surg. Pathol. 13(3), 202-206 (1989).
30 Aubert S, Wacrenier A, Leroy X et al. Weiss system revisited: a clinicopathologic and immunohistochemical study of 49 adrenocortical tumors. Am. J. Surg. Pathol. 26(12), 1612-1619 (2002).
31 Volante M, Buttigliero C, Greco E, Berruti A, Papotti M. Pathological and molecular features of adrenocortical carcinoma: an update. J. Clin. Pathol. 61(7), 787-793 (2008).
32 Soon PS, McDonald KL, Robinson BG, Sidhu SB. Molecular markers and the pathogenesis of adrenocortical cancer. Oncologist 13(5), 548-561 (2008).
33 Heppner C, Reincke M, Agarwal SK et al. MEN1 gene analysis in sporadic adrenocortical neoplasms. J. Clin. Endocrinol. Metab. 84(1), 216-219 (1999).
34 Kjellman M, Roshani L, Teh BT et al. Genotyping of adrenocortical tumors: very frequent deletions of the MEN1 locus in 11q13 and of a 1-centimorgan region in 2p16. J. Clin. Endocrinol. Metab. 84(2), 730-735 (1999).
35 Schulte KM, Mengel M, Heinze M et al. Complete sequencing and messenger ribonucleic acid expression analysis of the MEN I gene in adrenal cancer. J. Clin. Endocrinol. Metab. 85(1), 441-448 (2000).
36 Wachenfeld C, Beuschlein F, Zwermann O et al. Discerning malignancy in adrenocortical tumors: are molecular markers useful? Eur. J. Endocrinol. 145(3), 335-341 (2001).
37 Sidhu S, Marsh DJ, Theodosopoulos G et al. Comparative genomic hybridization analysis of adrenocortical tumors. J. Clin. Endocrinol. Metab. 87(7), 3467-3474 (2002).
38 Zhao J, Speel EJ, Muletta-Feurer S et al. Analysis of genomic alterations in sporadic adrenocortical lesions. Gain of chromosome 17 is an early event in adrenocortical tumorigenesis. Am. J. Pathol. 155(4), 1039-1045 (1999).
39 Giordano TJ, Thomas DG, Kuick R et al. Distinct transcriptional profiles of adrenocortical tumors uncovered by DNA microarray analysis. Am. J. Pathol. 162(2), 521-531 (2003).
40 de Fraipont F, El Atifi M, Cherradi N et al. Gene expression profiling of human adrenocortical tumors using complementary deoxyribonucleic acid microarrays identifies several candidate genes as markers of malignancy. J. Clin. Endocrinol. Metab. 90(3), 1819-1829 (2005).
·· Expression profile of candidate select genes in adrenocortical carcinoma were analyzed and prognostic markers identified.
41 Velazquez-Fernandez D, Laurell C, Geli J et al. Expression profiling of adrenocortical neoplasms suggests a molecular signature of malignancy. Surgery 138(6), 1087-1094 (2005).
42 Slater EP, Diehl SM, Langer P et al. Analysis by cDNA microarrays of gene expression patterns of human adrenocortical tumors. Eur. J. Endocrinol. 154(4), 587-598 (2006).
43 Lombardi CP, Raffaelli M, Pani G et al. Gene expression profiling of adrenal cortical tumors by cDNA macroarray analysis. Results of a preliminary study. Biomed. Pharmacother. 60(4), 186-190 (2006).
44 West AN, Neale GA, Pounds S et al. Gene expression profiling of childhood adrenocortical tumors. Cancer Res. 67(2), 600-608 (2007).
45 Fernandez-Ranvier GG, Weng J, Yeh RF et al. Identification of biomarkers of adrenocortical carcinoma using genomewide gene expression profiling. Arch. Surg. 143(9), 841-846; discussion 846 (2008).
· Evaluated a relatively large number of benign and malignant adrenocortical tumors with accurate diagnostic markers identified and confirmed by quantitative PCR.
46 Rogers S, Cambrosio A. Making a new technology work: the standardization and regulation of microarrays. Yale J. Biol. Med. 80(4), 165-178 (2007).
47 Steinhoff C, Vingron M. Normalization and quantification of differential expression in gene expression microarrays. Brief Bioinform. 7(2), 166-177 (2006).
48 Pollak MN, Schernhammer ES, Hankinson SE. Insulin-like growth factors and neoplasia. Nat. Rev. Cancer 4(7), 505-518 (2004).
49 Gicquel C, Bertagna X, Schneid H et al. Rearrangements at the 11p15 locus and overexpression of insulin-like growth factor-II gene in sporadic adrenocortical tumors. J. Clin. Endocrinol. Metab. 78(6), 1444-1453 (1994).
50 Gicquel C, Bertagna X, Gaston V et al. Molecular markers and long-term recurrences in a large cohort of patients with sporadic adrenocortical tumors. Cancer Res. 61(18), 6762-6767 (2001).
51 Almeida MQ, Fragoso MC, Lotfi CF et al. Expression of insulin-like growth factor-II and its receptor in pediatric and adult adrenocortical tumors. J. Clin. Endocrinol. Metab. 93(9), 3524-3531 (2008).
52 Barlaskar FM, Spalding AC, Heaton JH et al. Preclinical targeting of the type I insulin-like growth factor receptor in adrenocortical carcinoma. J. Clin. Endocrinol. Metab. 94(1), 204-212 (2009).
·· Comprehensive study demonstrating targeting of the IGF1 receptor in a preclinical model of adrenocortical carcinoma results in a significant antineoplastic effect.
53 Schmitt A, Saremaslani P, Schmid S et al. IGFII and MIB1 immunohistochemistry is helpful for the differentiation of benign from malignant adrenocortical tumours. Histopathology 49 (3), 298-307 (2006).
54 Fernandez-Ranvier GG, Weng J, Yeh RF et al. Candidate diagnostic markers and tumor suppressor genes for adrenocortical carcinoma by expression profile of genes on chromosome 11q13. World J. Surg. 32(5), 873-881 (2008).
55 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 survival. J. Clin. Oncol. 27(7), 1108-1115 (2009).
·· Large cohort of adrenocortical tumors that represent a clinical dilemma for diagnosis were distinct on gene-expression profile. The investigator also found gene-expression profile was an independent predictor of disease-free and overall survival in patients with adrenocortical carcinoma.
56 Giordano TJ, Kuick R, Else T et al. Molecular classification and prognostication of adrenocortical tumors by transcriptome profiling. Clin. Cancer Res. 15(2), 668-676 (2009).
·· Follow-up study where gene-expression profiling by prinicipal component analysis was an independent predictor of survival.
Affiliations
· Insoo Suh, MD University of California, San Francisco, Department of Surgery, Box 1674, San Francisco, CA 94143, USA
· Marlon A Guerrero, MD University of California, San Francisco, Department of Surgery, Box 1674, San Francisco, CA 94143, USA
· Electron Kebebew, MD University of California, San Francisco, Department of Surgery, Box 1674, San Francisco, CA 94143, USA Tel .: +1 415 885 3617 Fax: +1 415 885 7617 electron.kebebew@ucsfmedctr.org