ELSEVIER
Contents lists available at SciVerse ScienceDirect Molecular and Cellular Endocrinology
journal homepage: www.elsevier.com/locate/mce
Marcelse spd Celular Endocrinology
Review
Gene expression profiling in adrenocortical neoplasia
G. Assie a,b,c,d,*, T.J. Giordano %, J. Bertherat a,b,c,d,e,f
ª INSERM U1016, Institut Cochin, Paris, France
b CNRS UMR8104, Paris, France
” Université Paris Descartes, Sorbonne Paris Cité, Paris, France
d Department of Endocrinology, Assistance Publique Hôpitaux de Paris, Hôpital Cochin, Paris, France
e Center for Rare Adrenal Diseases, Assistance Publique Hôpitaux de Paris, Hôpital Cochin, Paris, France f Rare Adrenal Cancer Network COMETE-INCA, Paris, France
§ Department of Pathology, University of Michigan Health System, Ann Arbor, MI, USA
ARTICLE INFO
Article history: Received 31 July 2011
Received in revised form 14 September 2011 Accepted 14 September 2011 Available online 25 October 2011
Keywords:
Adrenocortical carcinoma Transcriptome
Diagnosis
Prognosis tumorigenesis
ABSTRACT
Transcriptome studies of adrenocortical tumors have shown clear differences between adenomas and car- cinomas and identified two subgroups of carcinomas with different prognoses. This review focuses on how transcriptomes have enriched our knowledge about genes previously identified by classical candidate gene approaches, uncovered novel genes relevant to adrenocortical tumor biology, helped to identify and under- stand specific pathway alterations, and advanced the overall translational relevance of this field of research. @ 2011 Elsevier Ireland Ltd. All rights reserved.
Contents
| 1. | Introduction | 111 |
| 2. | The candidate genes and the transcriptomes | 112 |
| 2.1. The high expression of IG2 in ACCs | 112 | |
| 2.2. TP53 is often altered in ACCs | 114 | |
| 2.3. Wnt-ßcatenin is often activated in ACCs | 114 | |
| 3. | Molecular classification of ACC | 114 |
| 3.1. The discrimination of ACCs and ACAs | 114 | |
| 3.2. The two types of ACCs | 114 | |
| 3.3. The three types of poor-prognosis ACCs | 115 | |
| 4. | Clinical developments | 115 |
| 4.1. Diagnosing a malignant neoplasm | 115 | |
| 4.2. Prognosis of ACC | 115 | |
| 4.3. Should we perform today a pan-genomic transcriptome array in adrenal cancer management? | 115 | |
| 5. | Perspectives | 116 |
| 5.1. The future place of transcriptome in the "omics" field | 116 | |
| 5.2. Moving from tumor groups to individual tumors characterization | 116 | |
| 5.3. Assess the clinical relevance of transcriptome | 116 | |
| 5.4. Impact of transcriptomes on treatments | 116 | |
| 6. | Conclusion | 116 |
| Acknowledgements | 116 | |
| References | 116 |
* Corresponding author. Address: Service d’endocrinologie, Hôpital Cochin, 27 rue du Fg Saint Jacques, 75014 Paris. Tel .: +33 158411820; fax: +33 158411805. E-mail address: guillaume.assie@cch.aphp.fr (G. Assie).
1. Introduction
Gene expression studies of cancers have been transformed by the advent of transcriptomes, transforming costly and time
consuming studies of individual candidate genes into a standard- ized procedure that yields reproducible pan-genomic expression information. Beyond confirming the role of candidate genes, the derivation of transcriptomes have brought major advances in understanding tumorigenesis, in tumor classification schemes, and in patients management(Quackenbush, 2006).
To date, about 10 primary studies have determined and re- ported the transcriptome of adrenocortical tumors (ACTs) (Giordano et al., 2003, 2009; de Fraipont et al., 2005; Velázquez-Fernández et al., 2005; Lombardi et al., 2006; Slater et al., 2006; West et al., 2007; Fernandez-Ranvier et al., 2008a, 2008b; Laurell et al., 2009; de Reyniès et al., 2009; Soon et al., 2009; Tömböl et al., 2009; Szabó et al., 2010). Consensual informa- tion has emerged, including a distinct difference in expression pro- files of adenomas (ACAs) and carcinomas (ACCs). Genes driving this difference include an enrichment in steroidogenesis-related genes in ACAs and an enrichment in cell-cycle related genes in ACCs. Re- cent reviews and a meta-analysis have summarized these aspects (Assié et al., 2010; Szabó et al., 2010; Ragazzon et al., 2011).
Tumor classifications based on the transcriptome also identified two groups of adult ACCs associated with different outcomes, as re- ported by three original studies (Giordano et al., 2009; Laurell et al., 2009; de Reyniès et al., 2009), and recently reviewed (Assié et al., 2010; Ragazzon et al., 2011). Of interest, the transcrip- tome-based prognosis remains significant after stratification by conventional pathological stage and grade.
In children, one study compared the transcriptome of ACCs to ACAs and normal adrenal, evoking similarities between the child- hood ACC transcriptome, and the foetal adrenal transcriptome, and between adulthood and childhood ACCs (West et al., 2007).
In this review, we will discuss how transcriptomes of adreno- cortical tumors have enriched our knowledge about genes previ- ously identified by classical candidate gene approaches, uncovered novel genes relevant to adrenocortical tumor biology,
helped to understand specific pathway alterations, and advanced the overall translational relevance of this research.
2. The candidate genes and the transcriptomes
Several genes have been identified for their relevance in adrenal carcinogenesis prior to the transcriptome era (Libé and Bertherat, 2005). Transcriptome studies have confirmed the differential expression of many of these genes and also identified their down- stream targets. Beyond a confirmatory role, transcriptome studies have also expanded our knowledge of these genes and provided an assessment of their relative importance to the biology of adre- nocortical neoplasia.
2.1. The high expression of IG2 in ACCs
ACCs occur in patients with Beckwith-Wiedmann syndrome (OMIM #130650). This syndrome is related to complex alterations at the 11p15 locus, resulting in the alteration of the expression of several genes in this region (see (Weksberg et al., 2003) for review) including IGF2 and H19. These two genes are regulated by a paren- tal imprint in a reciprocal manner. Beckwith-Wiedmann seems re- lated to a loss of the maternal IGF2 allele and expression of the paternal allele, resulting in an increased expression of IGF2 and re- duced expression of H19. Other genes, also at the 11p15 locus, are implicated in the Beckwith-Wiedmann syndrome, which include p57kip2, and KCNQ1. These two genes are also regulated by a parental imprint, but which is not related to the IGF2/H19 imprint. The interaction between these two imprinted regions is complex and incompletely understood.
Given the occurrence of ACCs in the Beckwith-Wiedmann syn- drome, expression of these genes at the 11p15 region was analyzed in sporadic ACCs well before genomic approaches were taken. High expression of IGF2 was found in approximately 90% of ACCs com-
D
N6
N10
A21
A28
A30
C4
C8
C11
C14
C17
C18
C19
36782_s_at
IGF2
Insulin-like growth factor 2 (somatomedin A)
1591_s_at
insulin-like growth factor 2 (somatomedin A)
1651_at 38116_at
KIAA0101
KIAA0101 gene product
34342_s_at
SPP1 SPP1
secreted phosphoprotein 1 (osteopontin, bone sialoprotein I, early T-lymphoc secreted phosphoprotein 1 (osteopontin, bone sialoprotein I, early T-lymphoc chromosome 20 open reading frame 1
39109_at
C20ORF1
2079_s_at
IGF2
insulin-like growth factor 2 (somatomedin A)
39542_at
ENC1
ectodermal-neural cortex (with BTB-like domain)
39230_at
DJ742C19.2 phorbolin (similar to apolipoprotein B mRNA editing protein)
Probe-Set
Gene Symbol
Unigene Title
N9
A22
H29
C7
C13
C15
C33
IGF2 UBCH10
ubiquitin carrier protein E2-C
2092_s_at
B
Paternal
TSSC3 SLC22A1L
CDKN1C KCNQ1
TSSC4 PHEMX
IGF2
H19
allele
KCNQ1OT1
Paternal allele
☒ Imprinted gene; the paternal allele is expressed
Oncogene
☐ Imprinted gene; the maternal allele is expressed
☐ Non imprinted gene
Tumor suppressor gene
Transcriptome-based tumor classification
Carcinomas (ACCs)
Adenomas (ACAs)
Poor Prognosis
Good Prognosis
p53-driven
ß-catenin-driven
Undetermined
Molecular events
IGF2 overexpression
Overexpression of Cell-cycle related genes
p53 alteration
Wnt/ßcatenin activation
DLG7-PINK1 prediction for non-metastatic tumors
IGF2 prediction for non-metastatic tumors
A
B
100
100
Disease-free survival (%)
80
predicted ACA
IGF2 low
Disease-free survival (%)
80
60
60
IGF2 high
40
40
predicted ACC
20
20
Log-rank p. : 1.06e-12
0
Log-rank p. : 2.84e-06
0
0
1
2
3
4
5
6
7
8
9
10
11
12
Years
0
1
2
3
4
5
6
7
8
9
10
11
12
Years
No. at Risk pred. ACA
No. at Risk
84 8
75
69
58
50
36
22
16
12
8
5
3
2
IGF2 low 77
69
62
50
42
29
16
10
8
5
3
2
1
bred. ACC
6
3
1
1
1
1
1
1
1
0
0
0
IGF2 high
15
12
10
9
9
8
7
7
5
4
2
1
1
Pathological prediction for non-metastatic tumors
C
100
Disease-free survival (%)
Weiss 0-2
80
60
40
Weiss 3-9
20
0
Log-rank p. : 4.75e-08
No. at Risk
0
1
2
3
4
5
6
7
8
9
10
11
12
Years
Weiss 0-2
71
63
58
49
43
30
17
11
9
7
4
Weiss 3-9
2
1
21
18
14
10
8
7
6
6
4
2
1
1
1
pared to ACAs (Gicquel et al., 1994). The first ACC transcriptome study and all subsequent ones have confirmed this result (Fig. 1). In children ACTs, IGF2 is also overexpressed in ACCs (West et al., 2007). It is therefore fair to ask, “What new knowledge has arisen from transcriptome analyses?”
First, as these studies have included almost all of expressed genes, it was easily demonstrated that altered IGF2 expression
was one of the most dominant trasncritpional events in ACCs, fur- ther underlying its importance in these tumors (Fig. 1).
Transcriptome data also provided a global picture of all the 11p15 genes expression in sporadic ACCs. A majority of tumors with a high IGF2 expression showed low H19 expression, which is in favor of the expression of the paternal allele and the loss of the maternal allele. Other imprinted genes in the 11p15 show
differential expression, also suggesting the loss of the maternal al- lele and/or imprinting (Fig. 1).
Transcriptomes studies have also shown that IGF2 is so uni- formaly and highly expressed that its expression was not responsi- ble for the ACC subclassification subsequently derived by unsupervised clustering (see below). This is ascertained by the clustering of the IGF2-negative ACCs with the IGF2-positives (Fig. 2).
From a clinical perspective, it is reasonable to ask what role can IGF2 play as a marker of a malignant neoplasm? Due to the occur- rence of around 10% of ACCs that lack high IGF2 expression, sensi- tivity of this marker is reduced. Without IGF2, the transcriptome performs better as a classification tool. By looking for a reduced number of genes deduced from the transcriptome, DLG7-PINK1 was found to perform better as a discrimination tool (Fig. 3).
Finally, IGF2 overexpression has led to new therapies designed to target this pathway. Transcriptome analyses could be used to as- sess the expression level of the IGF2, its receptors and binding pro- teins. Interestingly, IGF1-R, which mediates the main cellular effects of IGF2 in vivo (Gicquel and Le Bouc, 2006), is not downreg- ulated in ACCs overexpressing IGF2. Two preclinical studies have confirmed the implication of this pathway in ACC tumorigenesis (Almeida et al., 2008; Barlaskar et al., 2008), and an international clinical trial targeting the IGF1-R is ongoing (Clinical Trial.gov #NCT00924989).
2.2. TP53 is often altered in ACCs
TP53 is a major player in cancer. Attention was brought on TP53 in sporadic ACCs due to the occurrence of ACCs in Li-Fraumeni pa- tients, a syndrome often related to germline TP53 mutations (OMIM #151623). In adult sporadic ACCs, about one quarter of tu- mors harbor somatic TP53 mutations (Reincke et al., 1994; Barzon et al., 2001; Libè et al., 2007), and more than a half harbor loss of heterozygosity at the TP53 locus (Gicquel et al., 2001; Soon et al., 2008). Immunohistological alterations of p53 protein staining have also been identified. An unexpected high prevalence of unique exon 10 TP53 mutation in south Brazil was identified in childhood ACCs (Ribeiro et al., 2001).
Transcriptome studies have led to further understanding of the role of p53 in sporadic ACCs. Indeed, TP53 mutated tumors are en- riched in a subgroup of ACCs identified by unsupervised clustering of the tumors (Ragazzon et al., 2010). That means that specific transcriptome features are associated with this subgroup. Accord- ing to the transcriptome based classification, this subgroup lies within the group of aggressive ACCs (Fig. 2). Finally, genes posi- tively regulated by p53 such as RRM2B, TP53INP1 and MDM2, were found underexpressed in this subgroup. These data suggest that p53 is a driver of tumorigenesis in a specific subtype of high-grade ACCs, occurring relatively late in the tumor development.
2.3. Wnt-ßcatenin is often activated in ACCs
Wnt-ßcatenin is an important player in the adrenal develop- ment, starting early with the development of the urogenital ridge, further acting until the development of the normal cortex after birth (Keegan and Hammer, 2002). The pathway is activated by an accumulation of the ßcatenin protein in the cytoplasm, which translocates in the nucleus where it induces expression of several target genes. Aberrant Wnt-ßcatenin pathway activation plays a key role in the malignant transformation in several tissues, includ- ing the gut and liver.
Aberrant activation of the Wnt-ßcatenin pathway was identified in sporadic adrenocortical tumors (Tissier et al., 2005; Huang and He, 2008). Accumulation of ßcatenin in adrenocortical cells was
identified by immunohistochemistry, often related to activating Bcatenin mutations in almost all ACCs, and in one third of adenomas.
The role Wnt-Bcatenin activation was be further documented by the description of adrenocortical tumors in patients carrying APC mutations(Seki et al., 1992; Gaujoux et al., 2010).
Transcriptome studies have also provided important informa- tion about Wnt-ßcatenin. Indeed several Wnt-ßcatenin target genes are overexpressed in ACC, including baculoviral IAP repeat- containing 5 (BIRC5), ectodermal-neural cortex 1 (ENC1), pituitary tumor-transforming 1 (PTTG1), and twist homolog 1 (TWIST1) (de Fraipont et al., 2005; Velázquez-Fernández et al., 2005; Slater et al., 2006; Giordano et al., 2009; de Reyniès et al., 2009). More- over, gene set enrichment analyses showed enrichment in Wnt- ßcatenin genes in ACC compared to ACA (de Reyniès et al., 2009).
In addition, Wnt-ßcatenin alterations accumulate in a subgroup of aggressive ACCs also identified by transcriptome-based classifi- cation (Fig. 2). Positive transcriptional targets such as claudin 1 (CLDN1), axin 2 (AXIN2) and Leucine-rich repeat (LRR)-containing G-protein coupled receptor 5 (LGR5) were also found overexpres- sed in this subgroup of ACCs (Ragazzon et al., 2010). These data suggest that the Wnt-ßcatenin activation is a driver molecular alteration in a subgroup of ACCs, occurring late in tumorigenesis, similar to p53 mutation. Interestingly, p53 and ß-catenin muta- tions in aggressive ACCs seem mutually exclusive, thus defining two distinct subgroups of high-grade ACCs (Fig. 2).
3. Molecular classification of ACC
3.1. The discrimination of ACCs and ACAs
The clear discrimination of ACCs from ACAs is the most univer- sal finding that emerged from adrenocortical tumors transcrip- tome studies. Separation of ACA from ACC in unsupervised transcriptome-based classifications is very robust (Fig. 2), with strong agreement with conventional histopathology. Considering disease-free survival, agreement is also strong with the transcrip- tome-based classification, with almost no recurrence nor metasta- ses occurring in the group corresponding to benign tumors (de Fraipont et al., 2005; Giordano et al., 2009; de Reyniès et al., 2009). Interestingly, in the subgroup of tumors corresponding to malignant tumors, not all tumors recur, raising two hypotheses: either these tumors do not possess malignant potential (i.e. are not malignant), or these tumors are malignant, but were effectively treated by surgery. The pathological features of malignancy (Weiss scores of 3 and more), and the high number of chromosomal alter- ations in CGH in these tumors (Assié et al., 2009) (personal data) strongly support the second hypothesis.
Discrimination between ACCs and ACAs in childhood ACTs, in contrast to adult ACCs, is often difficult and unreliable with stan- dard histopathologic approaches. The unsupervised clustering analysis performed by West et al. showed no clear difference be- tween ACCs and ACAs either, probably reflecting a biological con- tinuum between benign and malignant tumors in children (West et al., 2007). These pediatric tumors would potentially greatly ben- efit from additional integrative genomic study.
3.2. The two types of ACCs
One of the major findings from unsupervised transcriptome- based tumor classification, reported by three studies (Giordano et al., 2009; Laurell et al., 2009; de Reyniès et al., 2009), is the exis- tence of two distinct groups of ACCs (Fig. 2). Considering overall survival associated with these tumors, a major difference between the two subgroups could be demonstrated. These studies described increased malignancy-related pathological features in the sub-
group of poor prognosis. One of the studies showed a significantly higher number of mitoses in this subgroup (so called high-grade tumors) (Giordano et al., 2009). However, in none of these studies was histopathology alone able to fully discriminate between poor and good prognosis ACCs with the same accuracy as the transcriptome.
One question remains: do the poor prognosis ACCs evolve from the good prognosis ACCs over time? Or do these two groups corre- spond to two independent types of ACCs? Data to separate these two possible mechanistic outcomes are limited. Two lines of evi- dences are in favor of the “two-types” of ACCs hypothesis. Indeed, if the poor prognosis ACCs evolved from the good prognosis, one would have expected the genomic alterations to accumulate, and therefore to be more abundant in tumors of the poor prognosis subgroup, compared to tumors of the good prognosis. In fact the number of chromosomal alterations is not increased in the sub- group of poor prognosis defined by the transcriptome (personal data). In addition, the transcriptome-based survival prediction re- mains significant after stratification on tumor extension, which means that the transcriptome-based subgroups do not only reflect different tumor stages. On the other hand, the increasing observa- tion of high-grade clones within otherwise low-grade ACC sup- ports “one-type” hypothesis (personal communication, TJG). However more work still need to be done to settle this issue and it may be that both mechanistic pathways play a role in the devel- opment of poor prognosis ACCs.
3.3. The three types of poor-prognosis ACCs
A recent study characterized in the subgroup of ACCs of poor prognosis, a third level of classification (Ragazzon et al., 2010). Unsupervised clustering identified three subgroups of ACCs. Interestingly, one was enriched in p53 mutations, the second in Bcatenin cytonuclear alterations and/or mutations, and the third presented neither of these two alterations (Fig. 2), illustrating that the full mutational spectrum in ACC is not known.
4. Clinical developments
4.1. Diagnosing a malignant neoplasm
The ACT transcriptome contains a strong signature of malignant behavior. This signature can obviously be translated to clinical practice. In one study, the pan-genomic transcriptome information was summarized to a minimal number of genes, actually two (DLG7 and PINK1) (de Reyniès et al., 2009). Assessment of their expression level by retrotranscriptase quantitative PCR was dem- onstrated to have a powerful diagnostic value in adrenocortical tu- mors. Of note, this tool was designed on disease-free survival, a stringent definition of malignancy. Indeed this definition could overcome the limitations of histopathology. From a statistical point of view, considering disease-free survival prediction was the only way to compare the predictor to pathology. In addition, disease- free survival prediction permitted to prove that the molecular pre- diction contained some information that was independent from the pathology. One drawback of this strategy is the classification as “at low risk of recurrence” of obviously malignant tumors, but that are completely resectable by surgery, as discussed in the pre- vious paragraph.
Is there a need for molecular malignancy tools? Certainly not for a majority of situations in which tumor classification into be- nign and malignant categories is straightforward using conven- tional histopathology. However two points deserve consideration. First the malignancy status of a tumor can be uncertain or undeter- mined. In the Weiss score system, these tumors correspond to
scores of 2 and 3 (Weiss, 1984; Weiss et al., 1989). The DLG7-PINK1 gene predictor of recurrence seems to perform better than the Weiss score, whatever the cut-off (Fig. 3B). Second, due to the rar- ity of adrenal cancer, not all pathologists are expert in adrenocor- tical pathology. Transcriptome studies are published by expert teams with highly skilled pathologists, who might not reflect the true level of all pathologists facing an adrenal specimen. Therefore, the performance of pathology for assessing malignancy might be overestimated in the published transcriptome studies.
In childhood ACTs, a set of genes differentially expressed be- tween ACCs and ACTs could be identified in one study (West et al., 2007). Further validation of the diagnostic value is expected.
4.2. Prognosis of ACC
The ACC transcriptome contains some strong prognostic infor- mation, as discussed in the previous paragraph. As for the predic- tion of recurrence, a prediction of the specific survival was proposed, based on summarized transcriptome information. This approach led to the identification of a marker based on the expres- sion level analysis of two genes. These two genes (BUB1B-PINK1) provide a strong prognostic assessment that remains significant after stratification on the tumor stage (de Reyniès et al., 2009).
Prognostic assessment of ACCs is a critical issue in clinical prac- tice. Tumor stage provides the strongest prognostic prediction. However, among each tumor stage, survival varies substantially. For metastatic disease, variable survival has been described, and prognostic parameters have been identified, reflecting the tumor grade (mitotic count in the primary) and the tumor extension (number of metastatic organs) (Assié et al., 2007). The progression slope provides a strong prognostic marker, but that can only be determined retrospectively. Determination of the prognosis might direct the treatments in the metastatic setting. In situations of good prognosis, especially in case of limited tumor extension, loco- regional approaches (mainly repeated surgery) might be proposed and be effective. In situations of poor prognosis, systemic treat- ments would be prioritized.
For ACCs limited to the adrenal gland and in apparent complete remission after an appropriate surgery, prognostic determination is even more critical. The tumor grade, based on mitotic count, seems an important prognostic factor (Weiss et al., 1989). However most experienced centers at present consider prognostic prediction in this case to be difficult. A molecular prediction in this setting would be of great potential. Finally, future studies on adjuvant therapy would probably benefit from appropriate tumor stratification.
4.3. Should we perform today a pan-genomic transcriptome array in adrenal cancer management?
Beyond scientific benefits, several lines of evidence would sup- port immediate clinical benefits of prospectively determining the transcriptome of each tumor that is not a straightforward adenoma.
Indeed, diagnostic information could be harvested, including the malignancy status and a confirmation of adrenocortical origin (especially in non-secreting tumors). Further, valuable prognostic information would also be obtained. In addition, costs has come down and the technology is widely available at academic centers. Finally in some cancer types, such as breast carcinoma (Kaufmann and Pusztai, 2011), transcriptome-based molecular typing has been proposed and is currently available.
Then why do not we do it? Several obstacles remain: merging transcriptome information from different experiments, even when derived from the same platform, is challenging (Xu and Wong, 2010). The best transcriptomic data is derived from unfixed, frozen
tumor tissue, which is often not obtained at the time of surgery. In addition, bioinformatics analyses are complex and require compar- ison to a large data set, ideally generated locally. Finally consider- ing common cancers, despite some available molecular tools, their use is currently limited, and no clear recommendation is available for their widespread use (Kaufmann and Pusztai, 2011).
5. Perspectives
5.1. The future place of transcriptome in the “omics” field
The field of “omics” is at a turn now, with the current rapid acceptance of high-throughput deep sequencing. However, the transcriptome should remain central, because it is an intermediate phenotype of biological features. Genome-related techniques do not have such a close and direct relationship to biology.
Transcriptomes in ACTs should improve: exon specific arrays al- ready exist, and deep sequencing should also be apply to transcrip- tome studies, with techniques like RNA sequencing, enabling extensive deciphering of alternative transcription (Hallegger et al., 2010).
5.2. Moving from tumor groups to individual tumors characterization
Transcriptome studies were first designed to characterize ACC as a group of tumors with common features. Unsupervised tumor classification identified subgroups that no other technique had de- scribed. The granulation of this tumor classification has moved for- ward with the input of additional information, either molecular (TP53 and Wnt-ßcatenin pathway analyses) or pathological (tumor grade).
The full molecular explanation of individual tumors is a goal to achieve. Integration of different genomic approaches has begun, combining transcriptome and tumor genome alterations (Szabó et al., 2010), or transcriptome and miRnome (Tömböl et al., 2009). Deep sequencing studies have now been launched in ACC, and resulting mutation information should yield important ad- vances in this field. Finally efforts should be pursued to integrate this information with other clinical features, such as pathology, hormone secretion, or any other relevant clinicopathologic information.
5.3. Assess the clinical relevance of transcriptome
Several challenges remain before the transcriptome and its de- rived tools can be fully employed in a routine laboratory setting. First, further validation is required. Prospective and retrospective validation of the molecular prognosis/predictive biomarkers are ongoing. Second, RNA handling is challenging in clinical practice, and robustness of the methods and/or technological improvements are needed. Finally ACC is a rare disease. Large cohorts can only be obtained in multicenter studies. International research networks such as the European Network For The Study of Adrenal Tumors (ENSAT) should continue and expand.
5.4. Impact of transcriptomes on treatments
Transcriptome studies should impact treatment of ACC patients at different levels: stratification of trials into prognostic subgroups, pharmacogenomics (prediction of response to a given chemother- apy), and identification of new molecular targets through a better understanding of the molecular mechanisms leading to tumorigen- esis in ACC. These topics remain to be approached by genomic studies in ACC.
6. Conclusion
Gene expression profiles of ACTs reflect tumor biology, pathol- ogy and clinical behavior. Further developments in basic science and ongoing translation of ACC molecular biology into novel molecular diagnostics are expected in the coming years.
Acknowledgements
The authors would like to acknowledge the “Ligue Contre le Cancer” for supporting our transcriptome studies through the “Carte d’identité des Tumeurs” project, especially Dr. Aurélien de Reynies and Dr. Jacqueline Godet.
Our genomic studies are supported in part by the Plan Hospita- lier de Recherche Clinique (AOM06179) to the COMETE Network, the Recherche Translationnelle DHOS/INCA 2009 (RTD09024) and the FP7 (ENSAT-Cancer) program.
References
Almeida, M.Q., Fragoso, M.C.B.V., Lotfi, C.F.P., Santos, M.G., Nishi, M.Y., Costa, M.H.S., Lerario, A.M., Maciel, C.C., Mattos, G.E., Jorge, A.A.L., Mendonca, B.B., Latronico, A.C., 2008. Expression of insulin-like growth factor-II and its receptor in pediatric and adult adrenocortical tumors. J. Clin. Endocrinol. Metab 93, 3524- 3531.
Assié, G., Antoni, G., Tissier, F., Caillou, B., Abiven, G., Gicquel, C., Leboulleux, S., Travagli, J .- P., Dromain, C., Bertagna, X., Bertherat, J., Schlumberger, M., Baudin, E., 2007. Prognostic parameters of metastatic adrenocortical carcinoma. J. Clin. Endocrinol. Metab 92, 148-154.
Assié, G., Raggazon, B., Reyniès, A. de, Carrasco, C., Guillot, M., Barreau, O., Rickmann, D., Tissier, F., Bertagna, X., Clauser, E., Bertherat, J., 2009. Characterization of Two Groups of Adrenocortical Carcinomas (ACCs) with Different Outcomes Revealed by Gene Profiling. Présenté à The Endocrine Society’s 91st Annual Meeting, Washington DC.
Assié, G., Guillaud-Bataille, M., Ragazzon, B., Bertagna, X., Bertherat, J., Clauser, E., 2010. The pathophysiology, diagnosis and prognosis of adrenocortical tumors revisited by transcriptome analyses. Trends Endocrinol. Metab. 21, 325-334.
Barlaskar, F.M., Spalding, A.C., Heaton, J.H., Kuick, R., Kim, A.C., Thomas, D.G., Giordano, T.J., Ben-Josef, E., Hammer, G.D., 2008. Preclinical targeting of the Type 1 Insulin-like growth factor receptor in adrenocortical carcinoma. J. Clin. Endocrinol. Metab. 78, 90-99.
Barzon, L., Chilosi, M., Fallo, F., Martignoni, G., Montagna, L., Palu, G., Boscaro, M., 2001. Molecular analysis of CDKN1C and TP53 in sporadic adrenal tumors. Eur. J. Endocrinol. 145, 207-212.
de Fraipont, F., El Atifi, M., Cherradi, N., Le Moigne, G., Defaye, G., Houlgatte, R., Bertherat, J., Bertagna, X., Plouin, P .- F., Baudin, E., Berger, F., Gicquel, C., Chabre, O., Feige, J .- J., 2005. 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, 1819- 1829.
de Reyniès, A., Assié, G., Rickman, D.S., Tissier, F., Groussin, L., Rene-Corail, F., Dousset, B., Bertagna, X., Clauser, E., Bertherat, J., 2009. Gene expression profiling reveals a new classification of adrenocortical tumors and identifies molecular predictors of malignancy and survival. J. Clin. Oncol. 8, 90-99.
Fernandez-Ranvier, G.G., Weng, J., Yeh, R .- F., Khanafshar, E., Suh, I., Barker, C., Duh, Q.Y., Clark, O.H., Kebebew, E., 2008a. Identification of biomarkers of adrenocortical carcinoma using genomewide gene expression profiling. Arch. Surg. 143, 841-846, discussion 846.
Fernandez-Ranvier, G.G., Weng, J., Yeh, R .- F., Shibru, D., Khafnashar, E., Chung, K .- W., Hwang, J., Duh, Q.Y., Clark, O.H., Kebebew, E., 2008b. Candidate diagnostic markers and tumor suppressor genes for adrenocortical carcinoma by expression profile of genes on chromosome 11q13. World J. Surg. 32, 873-881. Gaujoux, S., Pinson, S., Gimenez-Roqueplo, A .- P., Amar, L., Ragazzon, B., Launay, P., Meatchi, T., Libé, R., Bertagna, X., Audebourg, A., Zucman-Rossi, J., Tissier, F., Bertherat, J., 2010. Inactivation of the APC gene is constant in adrenocortical tumors from patients with familial adenomatous polyposis but not frequent in sporadic adrenocortical cancers. Clin. Cancer Res. 16, 5133-5141.
Gicquel, C., Le Bouc, Y., 2006. Hormonal regulation of fetal growth. Horm. Res. 65 (Suppl. 3), 28-33.
Gicquel, C., Bertagna, X., Schneid, H., Francillard-Leblond, M., Luton, J.P., Girard, F., Le Bouc, Y., 1994. Rearrangements at the 11p15 locus and overexpression of insulin-like growth factor-II gene in sporadic adrenocortical tumors. J. Clin. Endocrinol. Metab. 78, 1444-1453.
Gicquel, C., Bertagna, X., Gaston, V., Coste, J., Louvel, A., Baudin, E., Bertherat, J., Chapuis, Y., Duclos, J.M., Schlumberger, M., Plouin, P.F., Luton, J.P., Le Bouc, Y., 2001. Molecular markers and long-term recurrences in a large cohort of patients with sporadic adrenocortical tumors. Cancer Res. 61, 6762-6767.
Giordano, T.J., Thomas, D.G., Kuick, R., Lizyness, M., Misek, D.E., Smith, A.L., Sanders, D., Aljundi, R.T., Gauger, P.G., Thompson, N.W., Taylor, J.M.G., Hanash, S.M.,
2003. Distinct transcriptional profiles of adrenocortical tumors uncovered by DNA microarray analysis. Am. J. Pathol. 162, 521-531.
Giordano, T.J., Kuick, R., Else, T., Gauger, P.G., Vinco, M., Bauersfeld, J., Sanders, D., Thomas, D.G., Doherty, G., Hammer, G., 2009. Molecular classification and prognostication of adrenocortical tumors by transcriptome profiling. Clin. Cancer Res. 15, 668-676.
Hallegger, M., Llorian, M., Smith, C.W.J., 2010. Alternative splicing: global insights. FEBS J. 277, 856-866.
Huang, H., He, X., 2008. Wnt/beta-catenin signaling: new (and old) players and new insights. Curr. Opin. Cell Biol. 20, 119-125.
Kaufmann, M., Pusztai, L., 2011. Use of standard markers and incorporation of molecular markers into breast cancer therapy: consensus recommendations from an International Expert Panel. Cancer 117, 1575-1582.
Keegan, C.E., Hammer, G.D., 2002. Recent insights into organogenesis of the adrenal cortex. Trends Endocrinol. Metab. 13, 200-208.
Laurell, C., Velázquez-Fernández, D., Lindsten, K., Juhlin, C., Enberg, U., Geli, J., Höög, A., Kjellman, M., Lundeberg, J., Hamberger, B., Larsson, C., Nilsson, P., Bäckdahl, M., 2009. Transcriptional profiling enables molecular classification of adrenocortical tumours. Eur. J. Endocrinol. 161, 141-152.
Libé, R., Bertherat, J., 2005. Molecular genetics of adrenocortical tumours, from familial to sporadic diseases. Eur. J. Endocrinol. 153, 477-487.
Libè, R., Groussin, L., Tissier, F., Elie, C., René-Corail, F., Fratticci, A., Jullian, E., Beck- Peccoz, P., Bertagna, X., Gicquel, C., Bertherat, J., 2007. Somatic TP53 mutations are relatively rare among adrenocortical cancers with the frequent 17p13 loss of heterozygosity. Clin. Cancer Res. 13, 844-850.
Lombardi, C.P., Raffaelli, M., Pani, G., Maffione, A., Princi, P., Traini, E., Galeotti, T., Rossi, E.D., Fadda, G., Bellantone, R., 2006. Gene expression profiling of adrenal cortical tumors by cDNA macroarray analysis. Results of a preliminary study. Biomed. Pharmacother. 60, 186-190.
Quackenbush, J., 2006. Microarray analysis and tumor classification. N. Engl. J. Med. 354, 2463-2472.
Ragazzon, B., Libé, R., Gaujoux, S., Assié, G., Fratticci, A., Launay, P., Clauser, E., Bertagna, X., Tissier, F., de Reyniès, A., Bertherat, J., 2010. Transcriptome analysis reveals that p53 and {beta}-catenin alterations occur in a group of aggressive adrenocortical cancers. Cancer Res. 70, 8276-8281.
Ragazzon, B., Assié, G., Bertherat, J., 2011. Transcriptome analysis of adrenocortical cancers: from molecular classification to the identification of new treatments. Endocr. Relat. Cancer 18, R15-27.
Reincke, M., Karl, M., Travis, W.H., Mastorakos, G., Allolio, B., Linehan, H.M., Chrousos, G.P., 1994. P53 mutations in human adrenocortical neoplasms: immunohistochemical and molecular studies. J. Clin. Endocrinol. Metab. 78, 790-794.
Ribeiro, R.C., Sandrini, F., Figueiredo, B., Zambetti, G.P., Michalkiewicz, E., Lafferty, A.R., DeLacerda, L., Rabin, M., Cadwell, C., Sampaio, G., Cat, I., Stratakis, C.A., Sandrini, R., 2001. An inherited p53 mutation that contributes in a tissue- specific manner to pediatric adrenal cortical carcinoma. Proc. Natl. Acad. Sci. USA 98, 9330-9335.
Seki, M., Tanaka, K., Kikuchi-Yanoshita, R., Konishi, M., Fukunari, H., Iwama, T., Miyaki, M., 1992. Loss of normal allele of the APC gene in an adrenocortical carcinoma from a patient with familial adenomatous polyposis. Hum. Genet. 89, 298-300.
Slater, E.P., Diehl, S.M., Langer, P., Samans, B., Ramaswamy, A., Zielke, A., Bartsch, D.K., 2006. Analysis by cDNA microarrays of gene expression patterns of human adrenocortical tumors. Eur. J. Endocrinol. 154, 587-598.
Soon, P.S.H., Libe, R., Benn, D.E., Gill, A., Shaw, J., Sywak, M.S., Groussin, L., Bertagna, X., Gicquel, C., Bertherat, J., McDonald, K.L., Sidhu, S.B., Robinson, B.G., 2008. Loss of heterozygosity of 17p13, with possible involvement of ACADVL and ALOX15B, in the pathogenesis of adrenocortical tumors. Ann. Surg. 247, 157-164.
Soon, P.S.H., Gill, A.J., Benn, D.E., Clarkson, A., Robinson, B.G., McDonald, K.L., Sidhu, S.B., 2009. Microarray gene expression and immunohistochemistry analyses of adrenocortical tumors identify IGF2 and Ki-67 as useful in differentiating carcinomas from adenomas. Endocr. Relat. Cancer 16, 573-583.
Szabó, P.M., Tamási, V., Molnár, V., Andrásfalvy, M., Tömböl, Z., Farkas, R., Kövesdi, K., Patócs, A., Tóth, M., Szalai, C., Falus, A., Rácz, K., Igaz, P., 2010. Meta-analysis of adrenocortical tumour genomics data: novel pathogenic pathways revealed. Oncogene 29, 3163-3172.
Tissier, F., Cavard, C., Groussin, L., Perlemoine, K., Fumey, G., Hagneré, A .- M., Rene- Corail, F., Jullian, E., Gicquel, C., Bertagna, X., Vacher-Lavenu, M .- C., Perret, C., Bertherat, J., 2005. 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. 65, 7622-7627.
Tömböl, Z., Szabó, P.M., Molnár, V., Wiener, Z., Tölgyesi, G., Horányi, J., Riesz, P., Reismann, P., Patócs, A., Likó, I., Gaillard, R .- C., Falus, A., Rácz, K., Igaz, P., 2009. Integrative molecular bioinformatics study of human adrenocortical tumors: microRNA, tissue-specific target prediction, and pathway analysis. Endocr. Relat. Cancer 16, 895-906.
Velázquez-Fernández, D., Laurell, C., Geli, J., Höög, A., Odeberg, J., Kjellman, M., Lundeberg, J., Hamberger, B., Nilsson, P., Backdahl, M., 2005. Expression profiling of adrenocortical neoplasms suggests a molecular signature of malignancy. Surgery 138, 1087-1094.
Weiss, L.M., 1984. Comparative histologic study of 43 metastasizing and nonmetastasizing adrenocortical tumors. Am. J. Surg. Pathol. 8, 163-169.
Weiss, L.M., Medeiros, L.J., Vickery, A.L., 1989. Pathologic features of prognostic significance in adrenocortical carcinoma. Am. J. Surg. Pathol. 13, 202-206.
Weksberg, R., Smith, A.C., Squire, J., Sadowski, P., 2003. Beckwith-Wiedemann syndrome demonstrates a role for epigenetic control of normal development. Hum. Mol. Genet 12 Spec No 1, R61-8.
West, A.N., Neale, G.A., Pounds, S., Figueredo, B.C., Rodriguez Galindo, C., Pianovski, M.A.D., Oliveira Filho, A.G., Malkin, D., Lalli, E., Ribeiro, R., Zambetti, G.P., 2007. Gene expression profiling of childhood adrenocortical tumors. Cancer Res. 67, 600-608.
Xu, J .- Z., Wong, C .- W., 2010. Hunting for robust gene signature from cancer profiling data: sources of variability, different interpretations, and recent methodological developments. Cancer Lett. 296, 9-16.