DNA copy number analysis of Grade II–III and Grade IV gliomas reveals differences in molecular ontogeny including chromothripsis associated with IDH mutation status
- Adam Cohen†1Email author,
- Mariko Sato†2,
- Kenneth Aldape3,
- Clinton C. Mason1,
- Kristin Alfaro-Munoz3,
- Lindsey Heathcock3,
- Sarah T. South4, 6,
- Lisa M. Abegglen1,
- Joshua D. Schiffman†1, 5 and
- Howard Colman†1
© Cohen et al. 2015
Received: 12 May 2015
Accepted: 13 May 2015
Published: 20 June 2015
Isocitrate dehydrogenase (IDH) mutation status and grade define subgroups of diffuse gliomas differing based on age, tumor location, presentation, and prognosis. While some biologic differences between IDH mutated (IDH mut) and wild-type (IDH wt) gliomas are clear, the distinct alterations associated with progression of the two subtypes to glioblastoma (GBM, Grade IV) have not been well described. We analyzed copy number alterations (CNAs) across grades (Grade II–III and GBM) in both IDH mut and IDH wt infiltrating gliomas using molecular inversion probe arrays.
Ninety four patient samples were divided into four groups: Grade II–III IDH wt (n = 17), Grade II–III IDH mut (n = 28), GBM IDH wt (n = 25), and GBM IDH mut (n = 24). We validated prior observations that IDH wt GBM have a high frequency of chromosome 7 gain (including EGFR) and chromosome 10 loss (including PTEN) compared with IDH mut GBM. Hierarchical clustering of IDH mut gliomas demonstrated distinct CNA patterns distinguishing lower grade gliomas versus GBM. However, similar hierarchical clustering of IDH wt gliomas demonstrated no CNA distinction between lower grade glioma and GBM. Functional analyses showed that IDH wt gliomas had more chromosome gains in regions containing receptor tyrosine kinase pathways. In contrast, IDH mut gliomas more commonly demonstrated amplification of cyclins and cyclin dependent kinase genes. One of the most common alterations associated with transformation of lower grade to GBM IDH mut gliomas was the loss of chromosomal regions surrounding PTEN. IDH mut GBM tumors demonstrated significantly higher levels of overall CNAs compared to lower grade IDH mut tumors and all grades of IDH wt tumors, and IDH mut GBMs also demonstrated significant increase in incidence of chromothripsis.
Taken together, these analyses demonstrate distinct molecular ontogeny between IDH wt and IDH mut gliomas. Our data also support the novel findings that malignant progression of IDH mut gliomas to GBM involves increased genomic instability and genomic catastrophe, while IDH wt lower grade tumors are virtually identical to GBMs at the level of DNA copy number alterations.
Gliomas are the most frequent primary malignant brain tumors with an annual incidence of approximately 20,000 cases in the United States . Glioblastoma (GBM) is the most common glioma and remains nearly uniformly fatal, with a median survival under 16 months in aggressively treated patients . While these tumors are currently diagnosed by histopathology alone and generally treated based on histology and grade, recent findings identifying distinct molecular subgroups within these tumor types strongly suggest that improving treatments and patient survival will require detailed understanding of the biological and clinical differences between these subgroups.
Histopathologically, diffuse gliomas are categorized according to the WHO by histology (Astrocytoma, Oligodendroglioma, or Oligoastrocytoma) and grade [lower grade (grade II/III) versus glioblastoma (GBM, grade IV)]. Clinically, GBMs have been classified as primary or secondary on the basis of clinical presentation . Secondary GBMs, which are more common in young adults, display evidence of progression from a lower-grade tumor, whereas primary GBMs, which are more common in older adults, present as advanced cancers at diagnosis . Recently, large scale efforts have been made to identify the major genetic and epigenetic alterations and to define important molecular subtypes in GBM and lower grade gliomas [30, 43, 41]. The strongest prognostic factor for all glioma histologies is mutation in one of the isocitrate dehydrogenase genes (IDH1 or IDH2) , and mutation of these genes is seen at higher frequencies in lower grade gliomas and secondary GBMs.
Chromosome abnormalities in gliomas have been associated with various subgroups. A summary of over 400 GBMs showed gains in EGFR (7p12) in 30 %, GLI/CDK4/MDM2 (12q13-14) in 13 %, PIK3C2B/MDM4 (1q32) in 8 %, PDGFRA (4q12) in 8 %, and MET (7q31) in 4 % with deletions in CDKN2A/2B (9p21) in 47 %, PTEN (10q23) in 10 %, and RB1 (13q14) in 6 % . Primary GBMs commonly have gains in chromosome 7 and 19 and loss of chromosome 10 [31, 22, 25, 26, 46, 2]. Secondary or IDH mutated glioblastomas are less likely to have the above alterations and more likely to have gains in 8q and 10q accompanying simpler karyotypes [27, 22, 20, 25]. Grade II astrocytomas have been less well studied, with gains in 7q described in two studies [7, 12], and other alterations not replicated [7, 12, 46]. Although some groups have found worse prognosis in GBMs with either EGFR or chromosome 7 amplifications [15, 6, 26, 13], others, including the largest study (n = 532), found no association with outcome [14, 46, 13]. One reason for this inconsistency may be confounding due to the association of EGFR amplification with other known prognostic factors, such as age, G-CIMP status, or IDH mutation status [8, 3, 27, 22, 25, 20]. Only one study has looked at the prognosis of copy nuber alterations (CNAs) within subgroups defined by IDH status, suggesting that chromosome 7p gain and TP53 loss are prognostic in grade III gliomas with IDH mutation .
Gliomas have chromosomal instability, with a propensity for recurring patterns of CNAs [21, 36]. Although multiple mechanisms may be responsible for CNAs in gliomas, chromothripsis is a recently described form of localized CNA due to chromosomal catastrophe that may occur commonly in gliomas [18, 23]. Chromothripsis, which literally means “chromosome shattering,” can be identified from CNA technologies such as SNP microarrays. The association of chromothripsis with clinical factors and prognosis in gliomas has not been explored to date.
The purpose of this study is to determine CNAs within glioma subgroups defined by grade and IDH status. We deliberately chose to maximize the percent of IDH mut grade IV gliomas and IDH wt lower grade gliomas, as these are rare in most other studies. In addition, we examine prognostic CNAs within each glioma subgroup and chromothripsis as a function of grade and IDH status.
Materials and methods
Samples and nucleic acid extraction
We analyzed formalin-fixed, paraffin-embedded (FFPE) glioma specimens from 94 patients from M.D. Anderson Cancer Center (Houston, TX) and 20 autopsied normal brains (controls) from Huntsman Cancer Institute, University of Utah (Salt Lake City, UT). IRB approval was obtained from each institution. IDH mutation status was confirmed using direct sequencing . Gliomas were categorized by a single neuropathologist (KA) as either high grade (GBM) or lower grade (grade II or III). DNA was isolated using the Recoverall Total Nucleic Acid Isolation kit (AM1975, Applied Biosystems/Ambion, Austin, TX) and quantified with a high sensitivity, double strand specific, nucleic acid, fluorescent stain (PicoGreen, P7589, Invitrogen, Carlsbad, CA).
Copy number analysis
The DNA was plated in a 96-well plate with concentration goal of 7.5 ng/ul in a total volume of 40ul (300 ng total). The completed plates were sent to the Affymetrix Research Services Laboratory at Santa Clara, CA, and the OncoScan™ FFPE Express MIP assay was run as previously described [45, 35, 44]. The raw MIP data from the completed assay was loaded into Nexus Copy Number (BioDiscovery, Inc., El Segundo, CA). Stringency cutoffs for probe performance included call rates ≥90 % and standard deviations ≤0.3. BioDiscovery’s SNP-Rank Segmentation algorithm with Quadratic Wave Correction, a statistically based algorithm similar to Circular Binary Segmentation (CBS), was used to make copy number and loss of heterozygosity (LOH) calls . The significance threshold for segmentation was set at 5.0E-7 and required a minimum of five probes per segment. CNA thresholds were based on sample mosaicism, and set at 0.4 and −0.4 units of copy number from diploidy. High gain and homozygous loss were denoted by 1.2 and −1.2 units of copy number from diploidy. Genes were assigned to regions using the NCBI36/hg18 genome assembly on the UCSC genome browser. Gene gain was considered based on the copy number for that gene, without regard for entire chromosome gain or loss.
In order to assess the significance of the genomic alterations, Genomic Identification of Significant Targets in Cancer (GISTIC) was used to define deletions and gains and to calculate the q-value [40, 39], taking into account the frequency, amplitude and focality of the observed gains and deletions. CNAs with q < 0.25 were considered significant. Univariate and multivariate Cox proportional hazards models were fit to the data using Cox regression in SAS 9.3. Multivariable models were built using backward stepwise regression from a model including all variables with p < 0.1 in univariate analysis and maintaining variables with p < 0.05 in the multivariate analysis. Hazard ratios and p-values from the associated log-rank tests were reported. P-values for copy number alterations were adjusted using the Benjamini & Hochberg step-up false discovery rate (FDR) controlling procedure . The adjustment was done separately for each analysis. Hierarchical clustering was done using complete linkage disregarding the sex chromosomes. We also used gene ontology analysis to identify altered pathways within and between glioma subgroups using the ToppGene system . We disregarded pathways determined by genes grouped together on a single chromosome and, thus, affected as a single group by large CNAs. The heatmap in Fig. 3d was generated using the HeatMapImage module from Genepattern using the default color scheme . Fisher’s exact test was used for contingency table analyses.
Copy number alterations (CNAs) by subgroup
CNAs identified as significant within each of the four clinical/molecular subgroups using GISTIC q-values for CNAs are shown in Fig. 1b Due to their distinct chromosomal abnormalities and clinical characteristics, the lower grade oligodendroglial tumors with 1p/19q co-deletion (n = 5) were analyzed separately (Fig. 1a,c). A complete list of GISTIC significant CNAs in each subgroup defined by IDH status and grade is given in Additional file 1: Table S1.
IDH wt gliomas are similar regardless of grade
To examine molecular subgroups within tumors separated by IDH and 1p/19q status, we used unsupervised hierarchical clustering. In IDH wt gliomas, lower grade and grade IV gliomas clustered together in one large top level cluster (Fig. 2b), indicating that lower and high grade IDH wt tumors share similar CNA alterations. The CNA alterations seen most frequently across all grades of IDH wt gliomas were broad gain of chromosome 7 and loss of chromosome 10 (Arrows, Fig. 2b). This pattern of large CNAs in IDH wt gliomas contrasts with IDH mut gliomas, in which changes on chromosomes 7 and 10 were either absent or more focal around the EGFR, MGMT, and/or PTEN genes (Arrows, Fig. 2c).
Using functional gene ontology analysis to identify relevant pathways associated with significant CNAs, we found that IDH wt lower grade gliomas were enriched for alterations in pathways involving RB/checkpoint signaling, kinase binding, PI3K/AKT signaling, and cell cycle control. We also identified the pathways enriched in CNAs that differed between lower and high grade IDH wt gliomas. These pathways included base excision repair, telomerase extension, nucleotide excision repair, and repair of abasic sites, suggesting a small window of sensitivity may exist to DNA damaging agents early in IDH wt GBM development.
Progression to grade IV in IDH mut gliomas involves losses on chromosome 10 and increased chromosome instability
Among 1p/19q non-co-deleted IDH mut gliomas, unsupervised clustering identified two major clusters with significantly different percent of high and lower grade gliomas in each cluster (p = 0.018). One large cluster included 83 % (20) of the IDH mut GBMs but only 48 % (11) of the lower grade IDH mut gliomas. The other predominant cluster contained 35 % (8) of the lower grade IDH mut gliomas and 12 % (3) of the IDH mut GBMs. A third smaller cluster contained one GBM and four lower grade gliomas (Fig. 2c). The most significant difference (P = 5 × 10−5) between the two largest clusters was loss of the terminal end of the q arm of chromosome 10 including MGMT, which occurred in 80 % of the cluster with most of the GBMs and 9 % of the cluster with primarily lower grade gliomas (Additional file 4: Table S4). Loss of PTEN, which is more proximal on chromosome 10, was also associated with the two largest clusters, although not as tightly. Thus, it is not clear if the important gene on chromosome 10 is PTEN or MGMT or both.
Grade IV IDHmut gliomas are considered to be secondary GBMs that have progressed from lower grade gliomas. Therefore, differences between lower grade and grade IV IDH mut gliomas may indicate genes or pathways that are important for progression of IDHmut gliomas. In addition to the losses in 10q indicated above, grade IV IDH mut gliomas were more likely to have gains of 1q25.3 (SMG7, NCF2), 1q32.1 (KIF14, DDX59, BTG2), 6p21.1 (HSP90AB1 and other genes) and loss of 3p21 (multiple genes). A broad loss of heterozygosity (LOH) on 11p15 was also more common in the grade IV gliomas (Additional file 5: Table S5 and Fig. 3b). Applying functional gene ontology analysis to genes on these chromosome segments, the only enriched pathway was nitrogen compound transport (Additional file 6: Table S6). Both lower grade and grade IV IDH mut gliomas were enriched foralterations in the PI3K/AKT pathway. However, only IDH mut grade IV gliomas were enriched for alterations in pathways involving regulation of actin cytoskeleton, RAS, and EGFR. These differences suggest that RAS signaling and cytoskeletal abnormalities may play a role in progression of IDH mut gliomas.
Increased genomic instability is observed in IDH mut gliomas
We also examined the TCGA GBM and lower grade glioma datasets for total number of copy number alterations. Significantly more copy number alterations were seen in both the IDH mut Grade IV (mean 132.1, median 105 per sample) and IDH wt Grade IV (mean 132.8, median 96.5 per sample) tumors compared to the lower grade IDH mut (mean 63.04, median 53 per sample) and IDH wt (mean 53.9, median 40 per sample) gliomas. However, due to the small number of IDH mut GBM with copy number data (17), the power for comparing the number of CNA between IDH mut and IDH wt GBM was low. Moreover, unlike our samples, the lower grade and grade IV samples in the TCGA were run separately, so batch effects are possible.
Given the subgroup differences in CNA frequency, we examined the specific patterns of alterations across the whole genome and looked within groups at the specific CNAs on chromosomes with a high number of alterations. The term chromothripsis describes situations in which there are a large number of chromosomal rearrangements over localized chromosomal regions [10, 38]. In our analysis, we used the definition of at least 10 switches between two copy-number states (gain and loss) on at least one individual chromosome for a tumor to be considered to have chromothripsis. An example of a chromosome with chromothripsis is shown in Fig. 4b. By this definition, 11 of our samples contained chromothripsis. Chromothripsis was significantly more common in IDH mut Grade IV tumors than IDH wt (p = 0.002) or lower grade IDH mut (p = 0.05, Fig. 4c).
We hypothesized that loss of function of p53 would predispose to chromothripsis because of the inability of p53 deficient cells to undergo apoptosis in the face of chromosome shattering. Indeed, gliomas with chromosome loss at the TP53 locus or LOH at the TP53 locus were more likely to have chromothripsis than those with no alteration of TP53 (Fig. 4d), although this relationship was limited to Grade IV tumors.
The prognostic significance of chromothripsis is unknown. In our cohort, chromothripsis was not prognostic.
Alterations in cancer associated genes reveal the biological differences between molecular subtypes
Gain of EGFR and loss of PTEN and CDKN2A occur together frequently in both IDH wt lower grade and grade IV gliomas (all three occurring together in 53 % and 40 %, respectively), with no significant differences of these alterations by grade. On the other hand, EGFR gain is significantly rarer overall in IDH mut gliomas (17 % grade II–III, 25 % grade IV) and CDKN2A loss is slightly lower (39 % grade II–III, 42 % grade IV) compared with IDH wt. Moreover, seeing all three alterations is very rare in IDH mut gliomas, only occurring in 4 % regardless of grade. We did observe significant differences in PTEN loss associated with grade in IDH mut tumors (17 % IDH mut grade II–III and 46 % IDH mut grade IV [p = 0.025]). These findings suggest that loss of PTEN or genes near it on chromosome 10q may be a key and unique factor associated with progression of IDH mut tumors to grade IV.
To examine the functional significance of the chromosome alterations seen in the different groups, we examined a predetermined list of genes in pathways previously shown to be altered and functionally important in gliomas, including receptor tyrosine kinases (RTK), phosphatidyl-inositol-3-kinase, NF-κB, P53, and cell cycle regulators (Fig. 5b). These genes were considered altered if they were in an extended region identified by GISTIC analysis as having a q-value <0.25. Although most gliomas show alterations in all of these pathways, the mechanism by which the pathways are altered can differ. IDH wt gliomas had significantly more chromosome alterations affecting RTK signaling than IDH mut gliomas. PI3K pathway activation also differed based on IDH status: upstream changes such as PTEN deletion or AKT gain were more common in IDH wt gliomas and MTOR gain was significantly less common (p = 2 × 10−7, 0.002, and 1 × 10−5, respectively). Such differences could have implications for application of multiple targeted treatments to these glioma subtypes. Among cell cycle regulators, IDH wt gliomas were significantly more likely to have CDK1 loss and less likely to have cyclin A1 gene loss or cyclin D1 or E2 gene gain.
The strongest prognostic factors in the whole population were IDH status and grade (Fig. 1c). In multivariate analysis, the other significant variables were loss of the estrogen receptor B (ESR2), gain of CDKN1C, and TP53 loss, each of which was a negative prognostic factor (Additional file 7: Table S7). Given the biologic and clinical differences between the four subgroups defined by IDH status and grade, we sought to identify distinct prognostic factors within each subgroup and within the entire IDH mut and IDH wt groups. Although univariate analysis identified distinct copy number alterations in each subgroup that were significantly associated with survival in our cohort, none were significantly associated with survival when we attempted to validate them using 433 GBM and 181 lower grade glioma samples from the TCGA obtained via Nexus premier.
We present one of the first comparative analyses of CNAs among glioma subgroups defined by WHO grade and IDH mutation status. Confirming prior observations, we observe significant chromosomal differences between IDH-mutant and IDH-wild type tumors. When analyzing subgroups by grade and mutation status, we find few significant copy number differences between IDH wt lower grade and IDH wt grade IV gliomas. These genomic similarities support the concept that despite their histologic appearance, biologically these lower grade IDH wt tumors are pre-glioblastomas with a median survival a mere 7 months longer than grade IV gliomas, with few long-term survivors . In contrast, among IDH mut tumors, clustering based on copy number demonstrates that lower grade and grade IV gliomas with IDH mutations are distinct biologic entities; they also have distinct prognosis. The progression of IDH mut gliomas from lower grade to grade IV involves multiple CNAs, particularly on chromosome 10q, affecting biologically relevant pathways including: activation of PI3K signaling through loss of PTEN and gain of mTOR, as well as activation of cell cycle signaling through gain of CDK4, CDK6, and cyclinE2. MGMT loss may play a role as well, consistent with the resistance of MGMT unmethylatedt gliomas to alkylating agents.
In comparison to IDH mut gliomas, IDH wt gliomas have greater activation of receptor tyrosine kinase signaling through EGFR gain, MET gain, and BRAF gain, in addition to increased gains in cell cycle activators and losses of cell cycle inhibitors compared to IDH mut gliomas. This is likely to be biologically relevant, as others have shown that the number of CNAs in the receptor tyrosine kinase pathway correlates with pathway activation measured by downstream kinase phosphorylation . Amplification of EGFR has been shown to separate GBM into distinct clusters [8, 26, 2]. Although IDH mutation status was not reported in these clustering papers, alterations seen in the non-EGFR amplified group, such as losses on chromosome 13, mirror those seen in our IDH mut glioblastomas. The lack of EGFR amplification in IDH mut glioblastoma was also seen in the TCGA samples . A similar pattern of chromosome gains and losses distinguished primary and secondary glioblastoma, which have a high rate of IDH mutations . These results validate our findings that growth factor receptor signaling, particularly in the EGFR pathway, differs between IDH mut and IDH wt gliomas.
We found an unexpectedly large number of intrachromosomal breakpoints, also known as chromothripsis, in our IDH mut GBM tumors. Upon closer inspection, we observed that chromothripsis is more likely when TP53 is altered through deletion and/or LOH, and others have found that astrocytes lacking p53 have more chromosome breaks  and medulloblastoma due to inherited TP53 mutations have increased chromothripsis . Different definitions of chromothripsis have been proposed in the literature, and although many of our samples meet the common definition of chromothripsis, they do not all fit into every definition of chromothripsis [38, 19]. Therefore, we cannot conclude whether the massive intrachromosomal instability seen in IDH mut gliomas in our samples occurs in one event (“true” chromothripsis) or in sequential events over time (severe chromosomal instability). Nevertheless, our data support that IDH mut high grade tumors contain the highest number of alternating, intrachromosomal breakpoints.
Our study strengths include the relatively large number of IDH mut GBM and IDH wt lower grade gliomas relative to other data sets (including TCGA), allowing us to better characterize these uncommon groups. In addition, we have been able to use relatively novel, high-resolution MIP technology to analyze archived FFPE tissue with associated clinical variables and mature outcome data. Weaknesses include the overall small size of the series, which means that conclusions, particularly about between group differences and within group prognostic factors, must be taken as hypothesis-generating.
In conclusion, we have shown that IDH and grade define four distinct groups of 1p/19q non-co-deleted gliomas determined by functionally important CNAs and unique prognostic factors. IDH wt lower grade gliomas and grade IV gliomas are closely related and driven by common and well known alterations including EGFR amplification and PTEN deletion, while IDH mut lower grade gliomas remain functionally distinct from grade IV gliomas. The transition of IDH mut lower grade gliomas to grade IV gliomas involves loss of PTEN and dysregulation of cell cycle regulators, in addition to an apparent higher frequency of chromosomal instability and/or chromothripsis.
This research was funded in part by a grant from the National Institutes of Health/National Cancer Institute (5P50 CA127001) to H.C. and K.A. and by the Huntsman Cancer Foundation (to H.C.). We acknowledge Dr. Soheil Shams and BioDiscovery, Inc. for their assistance in analyzing the genomic copy number data. J.D.S. holds the Edward B. Clark, MD Chair in Pediatric Research, and J.D.S. and C.C.M. are supported through the Primary Children’s Hospital (PCH) Pediatric Cancer Program, funded by the Intermountain Healthcare Foundation and the Primary Children’s Hospital Foundation. A.L.C. acknowledges the inspiration of Jessica Jennifer Cohen. None of the authors has a competing financial interest.
- Benjamini YHY (1995) Controlling the false discovery rate: a practical and poweful approach to multiple testing. Journal of the Royal Statistical Society Series B (Methodological) 57:289–300Google Scholar
- Bredel M, Bredel C, Juric D, Harsh GR, Vogel H, Recht LD, Sikic BI (2005) High-resolution genome-wide mapping of genetic alterations in human glial brain tumors. Cancer research 65:4088–4096. doi:10.1158/0008-5472.CAN-04-4229PubMedView ArticleGoogle Scholar
- Brennan C, Momota H, Hambardzumyan D, Ozawa T, Tandon A, Pedraza A, Holland E (2009) Glioblastoma subclasses can be defined by activity among signal transduction pathways and associated genomic alterations. PloS one 4, e7752. doi:10.1371/journal.pone.0007752PubMed CentralPubMedView ArticleGoogle Scholar
- Brennan CW, Verhaak RG, McKenna A, Campos B, Noushmehr H, Salama SR, Zheng S, Chakravarty D, Sanborn JZ, Berman SH, Beroukhim R, Bernard B, Wu CJ, Genovese G, Shmulevich I, Barnholtz-Sloan J, Zou L, Vegesna R, Shukla SA, Ciriello G, Yung WK, Zhang W, Sougnez C, Mikkelsen T, Aldape K, Bigner DD, Van Meir EG, Prados M, Sloan A, Black KL et al (2013) The somatic genomic landscape of glioblastoma. Cell 155:462–477. doi:10.1016/j.cell.2013.09.034PubMed CentralPubMedView ArticleGoogle Scholar
- Chen J, Bardes EE, Aronow BJ, Jegga AG (2009) ToppGene Suite for gene list enrichment analysis and candidate gene prioritization. Nucleic acids research 37:W305–W311. doi:10.1093/nar/gkp427PubMed CentralPubMedView ArticleGoogle Scholar
- Crespo I, Vital AL, Nieto AB, Rebelo O, Tao H, Lopes MC, Oliveira CR, French PJ, Orfao A, Tabernero MD (2011) Detailed characterization of alterations of chromosomes 7, 9, and 10 in glioblastomas as assessed by single-nucleotide polymorphism arrays. The Journal of molecular diagnostics : JMD 13:634–647. doi:10.1016/j.jmoldx.2011.06.003PubMed CentralPubMedView ArticleGoogle Scholar
- Dahlback HS, Gorunova L, Brandal P, Scheie D, Helseth E, Meling TR, Heim S (2011) Genomic aberrations in diffuse low-grade gliomas. Genes, chromosomes & cancer 50:409–420. doi:10.1002/gcc.20866View ArticleGoogle Scholar
- de Tayrac M, Etcheverry A, Aubry M, Saikali S, Hamlat A, Quillien V, Le Treut A, Galibert MD, Mosser J (2009) Integrative genome-wide analysis reveals a robust genomic glioblastoma signature associated with copy number driving changes in gene expression. Genes, chromosomes & cancer 48:55–68. doi:10.1002/gcc.20618View ArticleGoogle Scholar
- Dolecek T, Propp J, Stroup N, Kruchko C (2012) CBTRUS statistical report: primary brain and central nervous system tumors diagnosed in the United States in 2005–2009. Neuro-oncology 15:v1–v49View ArticleGoogle Scholar
- Forment JV, Kaidi A, Jackson SP (2012) Chromothripsis and cancer: causes and consequences of chromosome shattering. Nature reviews Cancer 12:663–670PubMedView ArticleGoogle Scholar
- Gorovets D, Kannan K, Shen R, Kastenhuber ER, Islamdoust N, Campos C, Pentsova E, Heguy A, Jhanwar SC, Mellinghoff IK, Chan TA, Huse JT (2012) IDH mutation and neuroglial developmental features define clinically distinct subclasses of lower grade diffuse astrocytic glioma. Clinical cancer research : an official journal of the American Association for Cancer Research 18:2490–2501. doi:10.1158/1078-0432.CCR-11-2977View ArticleGoogle Scholar
- Hirose Y, Aldape KD, Chang S, Lamborn K, Berger MS, Feuerstein BG (2003) Grade II astrocytomas are subgrouped by chromosome aberrations. Cancer genetics and cytogenetics 142:1–7PubMedView ArticleGoogle Scholar
- Hobbs J, Nikiforova MN, Fardo DW, Bortoluzzi S, Cieply K, Hamilton RL, Horbinski C (2012) Paradoxical relationship between the degree of EGFR amplification and outcome in glioblastomas. The American journal of surgical pathology 36:1186–1193. doi:10.1097/PAS.0b013e3182518e12PubMed CentralPubMedView ArticleGoogle Scholar
- Huncharek M, Kupelnick B (2000) Epidermal growth factor receptor gene amplification as a prognostic marker in glioblastoma multiforme: results of a meta-analysis. Oncology research 12:107–112PubMedGoogle Scholar
- Hurtt MR, Moossy J, Donovan-Peluso M, Locker J (1992) Amplification of epidermal growth factor receptor gene in gliomas: histopathology and prognosis. Journal of neuropathology and experimental neurology 51:84–90PubMedView ArticleGoogle Scholar
- Jeuken J, van den Broecke C, Gijsen S, Boots-Sprenger S, Wesseling P (2007) RAS/RAF pathway activation in gliomas: the result of copy number gains rather than activating mutations. Acta neuropathologica 114:121–133. doi:10.1007/s00401-007-0239-0PubMedView ArticleGoogle Scholar
- Johnson DR, Leeper HE, Uhm JH (2013) Glioblastoma survival in the United States improved after Food and Drug Administration approval of bevacizumab: a population-based analysis. Cancer. doi:10.1002/cncr.28259Google Scholar
- Kim TM, Xi R, Luquette LJ, Park RW, Johnson MD, Park PJ (2013) Functional genomic analysis of chromosomal aberrations in a compendium of 8000 cancer genomes. Genome research 23:217–227. doi:10.1101/gr.140301.112PubMed CentralPubMedView ArticleGoogle Scholar
- Korbel JO, Campbell PJ (2013) Criteria for inference of chromothripsis in cancer genomes. Cell 152:1226–1236. doi:10.1016/j.cell.2013.02.023PubMedView ArticleGoogle Scholar
- Lai A, Kharbanda S, Pope WB, Tran A, Solis OE, Peale F, Forrest WF, Pujara K, Carrillo JA, Pandita A, Ellingson BM, Bowers CW, Soriano RH, Schmidt NO, Mohan S, Yong WH, Seshagiri S, Modrusan Z, Jiang Z, Aldape KD, Mischel PS, Liau LM, Escovedo CJ, Chen W, Nghiemphu PL, James CD, Prados MD, Westphal M, Lamszus K, Cloughesy T et al (2011) Evidence for sequenced molecular evolution of IDH1 mutant glioblastoma from a distinct cell of origin. Journal of clinical oncology : official journal of the American Society of Clinical Oncology 29:4482–4490. doi:10.1200/JCO.2010.33.8715View ArticleGoogle Scholar
- Loeper S, Romeike BF, Heckmann N, Jung V, Henn W, Feiden W, Zang KD, Urbschat S (2001) Frequent mitotic errors in tumor cells of genetically micro-heterogeneous glioblastomas. Cytogenetics and cell genetics 94:1–8PubMedView ArticleGoogle Scholar
- Maher EA, Brennan C, Wen PY, Durso L, Ligon KL, Richardson A, Khatry D, Feng B, Sinha R, Louis DN, Quackenbush J, Black PM, Chin L, DePinho RA (2006) Marked genomic differences characterize primary and secondary glioblastoma subtypes and identify two distinct molecular and clinical secondary glioblastoma entities. Cancer research 66:11502–11513. doi:10.1158/0008-5472.CAN-06-2072PubMedView ArticleGoogle Scholar
- Malhotra A, Lindberg M, Faust GG, Leibowitz ML, Clark RA, Layer RM, Quinlan AR, Hall IM (2013) Breakpoint profiling of 64 cancer genomes reveals numerous complex rearrangements spawned by homology-independent mechanisms. Genome research 23:762–776. doi:10.1101/gr.143677.112PubMed CentralPubMedView ArticleGoogle Scholar
- Metellus P, Coulibaly B, Colin C, de Paula AM, Vasiljevic A, Taieb D, Barlier A, Boisselier B, Mokhtari K, Wang XW, Loundou A, Chapon F, Pineau S, Ouafik L, Chinot O, Figarella-Branger D (2010) Absence of IDH mutation identifies a novel radiologic and molecular subtype of WHO grade II gliomas with dismal prognosis. Acta neuropathologica 120:719–729. doi:10.1007/s00401-010-0777-8PubMedView ArticleGoogle Scholar
- Misra A, Pellarin M, Nigro J, Smirnov I, Moore D, Lamborn KR, Pinkel D, Albertson DG, Feuerstein BG (2005) Array comparative genomic hybridization identifies genetic subgroups in grade 4 human astrocytoma. Clinical cancer research : an official journal of the American Association for Cancer Research 11:2907–2918. doi:10.1158/1078-0432.CCR-04-0708View ArticleGoogle Scholar
- Nigro JM, Misra A, Zhang L, Smirnov I, Colman H, Griffin C, Ozburn N, Chen M, Pan E, Koul D, Yung WK, Feuerstein BG, Aldape KD (2005) Integrated array-comparative genomic hybridization and expression array profiles identify clinically relevant molecular subtypes of glioblastoma. Cancer Res 65:1678–1686. doi:10.1158/0008-5472.CAN-04-2921PubMedView ArticleGoogle Scholar
- Noushmehr H, Weisenberger DJ, Diefes K, Phillips HS, Pujara K, Berman BP, Pan F, Pelloski CE, Sulman EP, Bhat KP, Verhaak RG, Hoadley KA, Hayes DN, Perou CM, Schmidt HK, Ding L, Wilson RK, Van Den Berg D, Shen H, Bengtsson H, Neuvial P, Cope LM, Buckley J, Herman JG, Baylin SB, Laird PW, Aldape K (2010) Identification of a CpG island methylator phenotype that defines a distinct subgroup of glioma. Cancer cell 17:510–522. doi:10.1016/j.ccr.2010.03.017PubMed CentralPubMedView ArticleGoogle Scholar
- Ohgaki H, Dessen P, Jourde B, Horstmann S, Nishikawa T, Di Patre PL, Burkhard C, Schuler D, Probst-Hensch NM, Maiorka PC, Baeza N, Pisani P, Yonekawa Y, Yasargil MG, Lutolf UM, Kleihues P (2004) Genetic pathways to glioblastoma: a population-based study. Cancer research 64:6892–6899. doi:10.1158/0008-5472.CAN-04-1337PubMedView ArticleGoogle Scholar
- Olshen AB, Venkatraman ES, Lucito R, Wigler M (2004) Circular binary segmentation for the analysis of array-based DNA copy number data. Biostatistics 5:557–572. doi:10.1093/biostatistics/kxh008PubMedView ArticleGoogle Scholar
- Phillips HS, Kharbanda S, Chen R, Forrest WF, Soriano RH, Wu TD, Misra A, Nigro JM, Colman H, Soroceanu L, Williams PM, Modrusan Z, Feuerstein BG, Aldape K (2006) Molecular subclasses of high-grade glioma predict prognosis, delineate a pattern of disease progression, and resemble stages in neurogenesis. Cancer cell 9:157–173. doi:10.1016/j.ccr.2006.02.019PubMedView ArticleGoogle Scholar
- Rao SK, Edwards J, Joshi AD, Siu IM, Riggins GJ (2010) A survey of glioblastoma genomic amplifications and deletions. Journal of neuro-oncology 96:169–179. doi:10.1007/s11060-009-9959-4PubMedView ArticleGoogle Scholar
- Rausch T, Jones DT, Zapatka M, Stutz AM, Zichner T, Weischenfeldt J, Jager N, Remke M, Shih D, Northcott PA, Pfaff E, Tica J, Wang Q, Massimi L, Witt H, Bender S, Pleier S, Cin H, Hawkins C, Beck C, von Deimling A, Hans V, Brors B, Eils R, Scheurlen W, Blake J, Benes V, Kulozik AE, Witt O, Martin D et al (2012) Genome sequencing of pediatric medulloblastoma links catastrophic DNA rearrangements with TP53 mutations. Cell 148:59–71. doi:10.1016/j.cell.2011.12.013PubMed CentralPubMedView ArticleGoogle Scholar
- Reich M, Liefeld T, Gould J, Lerner J, Tamayo P, Mesirov JP (2006) GenePattern 2.0. Nature genetics 38:500–501. doi:10.1038/ng0506-500PubMedView ArticleGoogle Scholar
- Scherer HJ (1940) A critical review: the pathology of cerebral gliomas. J Neurol Psychiatry 3:147–177PubMed CentralPubMedView ArticleGoogle Scholar
- Schiffman JD, Wang Y, McPherson LA, Welch K, Zhang N, Davis R, Lacayo NJ, Dahl GV, Faham M, Ford JM, Ji HP (2009) Molecular inversion probes reveal patterns of 9p21 deletion and copy number aberrations in childhood leukemia. Cancer genetics and cytogenetics 193:9–18. doi:10.1016/j.cancergencyto.2009.03.005PubMed CentralPubMedView ArticleGoogle Scholar
- Shapiro JR, Yung WK, Shapiro WR (1981) Isolation, karyotype, and clonal growth of heterogeneous subpopulations of human malignant gliomas. Cancer research 41:2349–2359PubMedGoogle Scholar
- Shibahara I, Sonoda Y, Kanamori M, Saito R, Yamashita Y, Kumabe T, Watanabe M, Suzuki H, Kato S, Ishioka C, Tominaga T (2012) IDH1/2 gene status defines the prognosis and molecular profiles in patients with grade III gliomas. International journal of clinical oncology 17:551–561. doi:10.1007/s10147-011-0323-2PubMedView ArticleGoogle Scholar
- Stephens PJ, Greenman CD, Fu B, Yang F, Bignell GR, Mudie LJ, Pleasance ED, Lau KW, Beare D, Stebbings LA, McLaren S, Lin ML, McBride DJ, Varela I, Nik-Zainal S, Leroy C, Jia M, Menzies A, Butler AP, Teague JW, Quail MA, Burton J, Swerdlow H, Carter NP, Morsberger LA, Iacobuzio-Donahue C, Follows GA, Green AR, Flanagan AM, Stratton MR et al (2011) Massive genomic rearrangement acquired in a single catastrophic event during cancer development. Cell 144:27–40. doi:10.1016/j.cell.2010.11.055PubMed CentralPubMedView ArticleGoogle Scholar
- Storey JD, Tibshirani R (2003) Statistical methods for identifying differentially expressed genes in DNA microarrays. Methods Mol Biol 224:149–157PubMedGoogle Scholar
- Storey JD, Tibshirani R (2003) Statistical significance for genomewide studies. Proc Natl Acad Sci U S A 100:9440–9445. doi:10.1073/pnas.1530509100PubMed CentralPubMedView ArticleGoogle Scholar
- Sturm D, Witt H, Hovestadt V, Khuong-Quang DA, Jones DT, Konermann C, Pfaff E, Tonjes M, Sill M, Bender S, Kool M, Zapatka M, Becker N, Zucknick M, Hielscher T, Liu XY, Fontebasso AM, Ryzhova M, Albrecht S, Jacob K, Wolter M, Ebinger M, Schuhmann MU, van Meter T, Fruhwald MC, Hauch H, Pekrun A, Radlwimmer B, Niehues T, von Komorowski G et al (2012) Hotspot mutations in H3F3A and IDH1 define distinct epigenetic and biological subgroups of glioblastoma. Cancer cell 22:425–437. doi:10.1016/j.ccr.2012.08.024PubMedView ArticleGoogle Scholar
- van den Bent MJ, Hartmann C, Preusser M, Strobel T, Dubbink HJ, Kros JM, von Deimling A, Boisselier B, Sanson M, Halling KC, Diefes KL, Aldape K, Giannini C (2013) Interlaboratory comparison of IDH mutation detection. Journal of neuro-oncology 112:173–178. doi:10.1007/s11060-013-1056-zPubMedView ArticleGoogle Scholar
- Verhaak RG, Hoadley KA, Purdom E, Wang V, Qi Y, Wilkerson MD, Miller CR, Ding L, Golub T, Mesirov JP, Alexe G, Lawrence M, O’Kelly M, Tamayo P, Weir BA, Gabriel S, Winckler W, Gupta S, Jakkula L, Feiler HS, Hodgson JG, James CD, Sarkaria JN, Brennan C, Kahn A, Spellman PT, Wilson RK, Speed TP, Gray JW, Meyerson M et al (2010) Integrated genomic analysis identifies clinically relevant subtypes of glioblastoma characterized by abnormalities in PDGFRA, IDH1, EGFR, and NF1. Cancer cell 17:98–110. doi:10.1016/j.ccr.2009.12.020PubMed CentralPubMedView ArticleGoogle Scholar
- Wang Y, Cottman M, Schiffman JD (2012) Molecular inversion probes: a novel microarray technology and its application in cancer research. Cancer genetics 205:341–355. doi:10.1016/j.cancergen.2012.06.005PubMedView ArticleGoogle Scholar
- Wang Y, Moorhead M, Karlin-Neumann G, Falkowski M, Chen C, Siddiqui F, Davis RW, Willis TD, Faham M (2005) Allele quantification using molecular inversion probes (MIP). Nucleic acids research 33, e183. doi:10.1093/nar/gni177PubMed CentralPubMedView ArticleGoogle Scholar
- Wiltshire RN, Herndon JE II, Lloyd A, Friedman HS, Bigner DD, Bigner SH, McLendon RE (2004) Comparative genomic hybridization analysis of astrocytomas: prognostic and diagnostic implications. J Mol Diagn 6:166–179. doi:10.1016/S1525-1578(10)60507-7PubMed CentralPubMedView ArticleGoogle Scholar
- Yahanda AM, Bruner JM, Donehower LA, Morrison RS (1995) Astrocytes derived from p53-deficient mice provide a multistep in vitro model for development of malignant gliomas. Molecular and cellular biology 15:4249–4259PubMed CentralPubMedGoogle Scholar
- Yan H, Parsons DW, Jin G, McLendon R, Rasheed BA, Yuan W, Kos I, Batinic-Haberle I, Jones S, Riggins GJ, Friedman H, Friedman A, Reardon D, Herndon J, Kinzler KW, Velculescu VE, Vogelstein B, Bigner DD (2009) IDH1 and IDH2 mutations in gliomas. The New England journal of medicine 360:765–773. doi:10.1056/NEJMoa0808710PubMed CentralPubMedView ArticleGoogle Scholar
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.