• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of neuroncolAboutAuthor GuidelinesEditorial BoardNeuro-Oncology
Neuro Oncol. Dec 2008; 10(6): 995–1003.
PMCID: PMC2719013

Prevalence of copy-number neutral LOH in glioblastomas revealed by genomewide analysis of laser-microdissected tissues

Abstract

We have employed a laser-capture microdissection technique and single-nucleotide polymorphism arrays to characterize genomic alterations associated with the development of glioblastoma multiforme (GBM). Combined analysis of loss of heterozygosity (LOH) and copy number revealed that more than half (56.3%) of the 254 identified LOH loci showed no copy-number alteration, indicating the presence of copy-number neutral LOH (cnLOH). Furthermore, we found a GBM case that showed cnLOH in 18 of the 22 autosomes. These results were confirmed by quantitative real-time PCR, microsatellite analysis, and fluorescence in situ hybridization. The high rate of cnLOH suggests that epigenetic abnormalities of many genes are involved in the development and progression of GBMs.

Keywords: copy-number alteration, copy-number neutral LOH, glioblastoma, loss of heterozygosity, single-nucleotide polymorphism array

Despite recent advances in surgery, radiotherapy, and chemotherapy, the median survival time for patients with glioblastoma multiforme (GBM), the most malignant brain tumor, remains short, around 1 year after the initial surgery.1 Like many other cancers, GBM tumorigenesis is believed to involve a series of consecutive genetic alterations, and detailed genetic characterization of GBMs is expected to lead to improved diagnostic criteria, provide deeper insights into the mechanisms of tumorigenesis, and contribute to the possible development of novel therapeutic strategies. Conventional molecular genetic analyses of GBMs, such as comparative genomic hybridization (CGH) and fluorescence in situ hybridization (FISH), often reveal losses of chromosomes 9p, 10, 17p, 19q, and 22 and gains at loci on chromosomes 7 and 20.2,3 Inactivation of various tumor suppressor genes, including TP53, RB1, CDKN2A, and PTEN, and overexpression of oncogenes such as RASD1, EGFR, and AKT1 are associated with GBM tumorigenesis, as revealed by studies targeted to the candidate genes. However, the detailed mechanisms of tumorigenesis have been poorly understood.2,3

Loss of heterozygosity (LOH) analysis is an effective alternative approach to identify genetic aberrations in cancer. Conventionally, LOH has been detected by determining allelic imbalances using microsatellite markers. However, the resolution of this method is inevitably low because of the scarcity of the genome markers. We previously developed a single-nucleotide polymorphism (SNP)-based method in which PCR products are postlabeled with fluorescent dyes and analyzed in an automated capillary electrophoresis system under single-strand conformational polymorphism (SSCP) analysis conditions.4 This method allowed a high-resolution and high-sensitivity LOH detection, owing to the abundance of SNP markers and the highly quantitative nature of the SSCP analysis. Its sensitivity allowed the detection of allelic imbalances in surgically obtained specimens, which are often heavily contaminated with nontumorous tissues. However, the method is not suitable for genomewide analysis, because it requires a primer pair for each SNP marker, and the cost for the analysis using many markers can be formidable.

Recently, several algorithms have been developed to quantitatively interpret hybridization signals obtained from the analysis using high-density oligonucleotide microarrays, such as SNP arrays. This allowed the simultaneous detection of LOH and copy-number change in tumor samples at the SNP level in a genomewide fashion by single-platform analysis. The disadvantage of the SNP-array analysis is that the signal obtained for each hybridization is inevitably weak (sacrificed to attain high SNP density), and the resultant low signal-to-noise ratio does not allow confident detection of LOH or copy-number change for any tumor sample that is heavily contaminated with normal cells (30%–50% of the total).5,6 Because GBM tissues are often highly heterogeneous and infiltrated with normal cells, such as vascular endothelial cells, infiltrated leukocytes, or interstitial cells, selection of tumor cells before extraction of DNA is often critical for confident detection of tumor-specific genetic changes by SNP-array analysis. In the present study, we employed the laser-capture microdissection method to obtain DNA that was derived exclusively from a tumor. We examined both LOH and copy-number alterations using these samples and found that the majority of GBMs harbor LOH without copy-number alterations, which was designated as acquired uniparental disomy or copy-number neutral LOH (cnLOH). Although cnLOH has been detected in some other cancer cells using SNP microarray analysis,711 this is the first report describing cnLOH in GBMs to such an extent.

Materials and Methods

Sample Acquisition

GBM samples were obtained from 14 patients during surgery at Kyushu University Hospital or other affiliated institutions. A part of the tumor tissue was saved for histopathologic examination, and the remainder was snap-frozen in liquid nitrogen and stored at −80°C. Corresponding wild-type DNA was isolated from a blood sample from each patient using a QIAamp DNA Blood Mini Kit (Qiagen, Valencia, CA, USA). The present investigation was approved by the Ethics Committee of Kyushu University.

Laser-Capture Microdissection and DNA Extraction

Laser-capture microdissection was performed in all GBM samples to obtain a homogeneous population of tumor cells. Two 10-μm-thick frozen sections were cut from each GBM tissue: one section for hematoxylin and eosin staining to confirm the orientation of the tissue, and the other section for microdissection. Sections for microdissection were placed on PALM membrane glass slides (P.A.L.M. Microlaser Technologies AG, Bernried, Germany), briefly air-dried (30 sec), and fixed in 100% methanol on ice for 1 min. The sections were then stained with 0.05% toluidine blue for 30 sec at room temperature, washed briefly with distilled water, and air-dried. A PALM Robot Microbeam system (P.A.L.M.) was used for laser cutting and separation of selected tissue areas. The selected tumor areas were confirmed to be cleared from nontumorous cells (e.g., vascular endothelial cells, necrotic tissues, and normal astrocytes) (Fig. 1). We combined 50–100 captures for each specimen and transferred them into 0.5-ml tubes containing 180 μl of ATL buffer provided with the QIAamp DNA Mini Kit (Qiagen). DNA was extracted using the kit according to the manufacturer’s instructions. DNA quantity was determined using Quant-iT dsDNA Reagent Kit (Invitrogen, San Diego, CA, USA). On average, 1 μg of DNA was obtained from each tumor (Table 1).

Fig. 1.
Examples of glioblastoma cell microdissections. (A) Toluidine blue-stained section before microdissection. Tumor cells, bounded by green lines, were collected free from nontumorous tissues such as proliferative endothelial cells (arrows) and necrotic ...
Table 1.
Characteristics of study subjects

SNP-Array Analysis

GeneChip Mapping 50K Xba assay kit (Affymetrix Inc., Santa Clara, CA, USA) was used for the analysis. DNA samples were processed according to the standard protocol provided by the supplier. Briefly, 250 ng of genomic DNA was digested with XbaI, ligated to adapters, and amplified by polymerase chain reaction (PCR) using a single primer. After purification with MinElute 96 UF PCR Purification Kit (Qiagen), the amplified products were quantified, fragmented, labeled, and subsequently hybridized to the SNP arrays. After washing and staining, the arrays were scanned using a GeneChip Scanner 3000 (Affymetrix). Scanned data files were generated using Affymetrix GeneChip Operating Software, version 1.2, and genotype calls were made automatically by GeneChip DNA Analysis Software, version 3.0. Genotypes and probe intensities derived from blood and cancer DNA were loaded into the dChip software package (http://www.dchip.org/) for LOH and copy-number analysis, as previously described.6,12 Briefly, LOH call was assigned in a pair of tumor and matched blood as LOH or allelic imbalance (heterozygosity call in blood, homozygosity call in tumor), retention of heterozygosity (heterozygosity call in both blood and tumor), or noninformative (homozygosity call in blood). The hidden Markov model (HMM) from dChip was used to infer regional LOH with an LOH call threshold set at 0.5. For calculating the copy number, data were normalized to a baseline array with median signal intensity with the invariant set normalization method. The model-based (perfect match/mismatch) method was used to obtain the signal values after normalization.6 Normalized intensities in matched blood DNA samples were used as the reference set to calculate the copy number of each marker in tumor samples. The HMM was used to infer copy-number change along each chromosome.

Quantitative Real-Time PCR

Quantitative real-time PCR was performed on an ABI 7500 Fast System (Applied Biosystems, Foster City, CA, USA) using the Power SYBR Green PCR Master Mix (Applied Biosystems). PCR primers of six different genes were designed using Primer Express 3.0 software (Applied Biosystems) (Table 2). All reactions were performed in triplicate using the following PCR conditions: 95°C for 10 min followed by 40 cycles at 95°C for 15 sec and 60°C for 1 min. An additional cycle of 95°C for 15 sec, 60°C for 1 min, and gradual heating to 95°C was run at the end to measure the dissociation curve for quality control. Data evaluation was performed using the ABI 7500 Fast System Sequence Detection System software and RQ software (Applied Biosystems). Target gene copy number was determined by the relative quantitative comparative threshold cycle (ΔΔCt) method using the LINE-1 repetitive element as the endogenous control.13,14 In this method, the relative copy-number difference of the target gene in tumor sample against reference sample is given by 2ΔΔCt, where ΔΔCt is defined by the following equation:

ΔΔCt= (Ct of target gene in tumor sampleCt of LINE-1 in tumor sample)  (Ct of target gene in reference sample Ct of LINE-1 in reference sample)

Here, matched blood DNA was used as the reference sample and was assumed to have the copy number of 2.

Table 2.
Primer sequences for real-time PCR

Microsatellite Analysis

Tumor and matched blood DNA was evaluated by a PCR-based LOH assay using 48 microsatellite markers. PCR and fluorescence labeling were performed according to a previously described method.15 Capillary electrophoresis was performed with a 3730 Prism Genetic Analyzer (Applied Biosystems). Raw electrophoresis data were analyzed with GeneMapper analysis software (Applied Biosystems). Allelic status was assessed based on the criteria established in a previous study.15 Microsatellite instability was determined by the appearance of extra bands on tumor DNA examination.

Fluorescence In Situ Hybridization (FISH)

Paraffin sections 6–10 μm thick on slides were prewarmed to 60°C overnight and deparaffinized in two changes of xylene, followed by rinsing in 100% ethanol. The slides were immersed in 0.2 M HCl for 20 min at room temperature and then incubated in 1 M sodium thiocyanate for 20 min at 80°C. After cooling to room temperature, the slides were treated with protease solution containing 25 mg protease I (Vysis, Downers Grove, IL, USA) in 0.01 M HCl at 37°C for 30–60 min and then fixed in 10% neutral buffered formalin. After dehydration in an ethanol series (70%, 85%, and 100%), dual-probe hybridization was performed using probes that detected centromere regions of chromosomes 10 and 2 (Spectrum Green-labeled CEP 10 and Spectrum Orange-labeled CEP 2, respectively, obtained from Vysis), according to the manufacturer’s protocol. After incubation at 37°C for 18–24 h, the slides were washed once in 0.3% NP-40/2× saline-sodium citrate at room temperature for 5 min, followed by heating at 73°C for 2 min in the same solution. The slides were then damp dried, and the nuclei counterstained with DAPI-I solution (4,6-diamidino-2-phenylindole-I; Vysis). Two hundred nonoverlapping nuclei were counted under an Olympus BX50 fluorescence microscope equipped with multiple filters (Olympus Australia Pty. Ltd., Mount Waverley, VIC, Australia). A tumor sample was considered to be monosomic for the examined chromosome according to the criteria adopted by Cianciulli et al.16

Results

We performed a comprehensive analysis of chromosomal aberrations in 14 GBMs using Affymetrix Mapping 50K arrays, to achieve a genomewide evaluation of the copy number and LOH status. It is critical to obtain highly purified tumor DNA to examine GBM specimens, because they often contain prominent proliferating endothelial cells and massive necrotic tissues, occasionally involving neighboring cerebral cortex cells (Fig. 1A).17 Therefore, laser-capture microdissection was performed for all tumor samples to obtain genuine tumor cell populations (Fig. 1B). We obtained enough tumor DNA to perform SNP-array analysis, and favorable SNP call rates from tumor tissue (97.7 ± 0.90) and matched blood (98.6 ± 0.83) were obtained (Table 1).

The 50K SNPs provided good coverage of the whole genome at a sufficient resolution to enable identification of small LOH regions, as well as LOH involving the whole chromosome (Fig. 2). We defined the LOH region as two or more consecutive SNPs showing LOH calls. In total, we identified 254 LOH loci in 14 GBM cases, with a median of 18 LOH loci per patient (range, 0–60; Table 3). Interstitial deletion or gain of either whole or parts of chromosomal regions is a common mechanism of LOH in human cancer cells. We then estimated copy numbers for these 254 LOH loci and divided them into three categories: LOH with copy-number loss (≤1), LOH with neutral copy number (=2), and LOH with copy-number gain (≥3). Results indicated that more than half of the LOH loci (143 of 254, 56.3%) were copy-number neutral events. These cnLOH loci were observed in 12 of 14 cases (85.7%), although the proportion relative to the total LOH number varied widely. Also, the copy-number loss accounted for only 34.25% of LOH regions, and those with a copy-number gain accounted for 9.45% (Table 3). Although many small interstitial cnLOH regions scattered on various chromosomes were observed, complete cnLOH regions on specific chromosomes (chromosomes 2, 3, and 10) were evident in multiple cases (Fig. 2).

Fig. 2.
Genomewide loss of heterozygosity (LOH) patterns and copy-number changes in glioblastomas. (A) Chromosomes are joined and presented as indicated by the numbers on the left. The results of each sample, indicated by the number at the top, are presented ...
Table 3.
Loss of heterozygosity (LOH) regions and corresponding copy-number changes

To validate the copy-number data from the SNP array analysis, we determined the copy-number score of six different genes by quantitative real-time PCR in all tumor samples. The copy numbers between 0 and 3 estimated by the SNP-array experiments were confirmed with reasonable accuracy (0.24 ± 0.13, 0.85 ± 0.28, 1.99 ± 0.22, and 2.98 ± 0.31, respectively, R2 = 0.89) by real-time PCR. On the other hand, for the high copy-number regions, much higher values of copy-number gain were observed by the real-time PCR estimation than by the SNP-array estimation (Table 4). This is in agreement with the general notion that SNP-array-based analysis tends to seriously underestimate the copy number of highly amplified regions.

Table 4.
Comparison of copy-number estimations by single-nucleotide polymorphism (SNP) array and quantitative real-time PCR (qPCR) analyses

In both dChip and quantitative PCR analyses, the overall ploidy was assumed to be 2 for both the tumor and the normal samples. However, this assumption may not be true for a tumor with extreme aneuploidy. Therefore, we reevaluated the copy-number status by FISH experiments using centromere-specific probes. A typical case of LOH associated with copy-number loss is shown in Fig. 3B. In this sample, in good agreement with SNP-array-based LOH and copy-number data, in the majority of the cells only one hybridization signal was detected for the CEP 10 probe, whereas two signals were detected for the CEP 2 probe. Thus, in that sample, most of the tumor cells were likely to be in the G1 phase, and they were disomic for chromosome 2 while monosomic for chromosome 10 (Table 5). On the other hand, in samples 1 and 11, two signals were detected for both CEP 2 and CEP 10 probes in the majority of observed nuclei. Because LOH was observed for chromosome 10 of sample 11 and both chromosomes 2 and 10 in sample 1, we concluded that these chromosomes were cnLOH loci in the respective samples (Fig. 3A,C and Table 5).

Fig. 3.
Confirmation of copy-number neutral loss of heterozygosity (cnLOH) in glioblastoma multiforme samples. The results of single-nucleotide polymorphism (SNP) array and fluorescence in situ hybridization (FISH) analysis are compared for samples 1, 5, and ...
Table 5.
Summary of fluorescence in situ hybridization analysis

Discussion

Using array-based genotype and copy-number analysis, we have shown that LOH without copy-number reduction is a frequent event in GBM. This observation was further supported by quantitative PCR and FISH analyses. Recently, LOH without copy-number reduction has also been reported to be a common genetic event in other cancers.711,18,19 In pancreatic cancer, for instance, a study using an Affymetrix 100K array reported that 67.5% of all LOH loci in 26 pancreatic cancer cell lines showed copy-neutral or copy-gain LOH events, whereas the remaining 32.2% of LOH loci appeared to be associated with copy-number reduction.18 Another study using cervical cancer cell lines also demonstrated that approximately 75% of LOH events did not show copy-number change.19 Our data from the present study demonstrating that 56.3% of all LOH loci in analyzed GBMs were copy-number neutral are comparable to those studies on other cancers. To our knowledge, this is the first study demonstrating a high prevalence of cnLOH in GBMs.

cnLOH in tumors of the CNS was first described by James et al.,20 who reported that loss of allelic dosage (allelic imbalance) on chromosome 17 without copy-number reduction was often observed in low-grade astrocytomas. As for GBMs, although several LOH studies using SNP-array analysis have been reported, the copy number of the LOH sites has not been previously evaluated.21,22 Using the CGH technique, Wiltshire et al.23 analyzed 22 GBM cases and demonstrated that the proportion of LOH that did not indicate loss by CGH was 16% on chromosome 10 and 17% on chromosome 9. The proportion of LOH on chromosome 10 in GBMs evaluated by studies using microsatellite markers has been somewhat higher (70%–80%) than that of copy-number loss detected by CGH studies (50%–70%),2528 suggesting the discrepancy between CGH and LOH data might be due to cnLOH.24

In the present study, an extraordinary genotype was found in one case (sample 1), where 18 of 22 autosomes were LOH loci, all of them cnLOH (Fig. 2). We performed additional LOH analysis using microsatellite markers to confirm these unusual LOH patterns. As shown in Fig. 4, the results completely matched those of the SNP-based LOH. We identified no clinical peculiarities in this patient, but further study of possible correlation of this event with clinical manifestation (e.g., sensitivity to chemotherapy) in this or similar cases is worth pursuing.

Fig. 4.
Confirmation of loss of heterozygosity (LOH) in sample 1 using microsatellite analysis. For each chromosome, the left column shows results of PCR-based microsatellite analysis. Each marker is annotated as LOH (black), retention of heterozygosity (white), ...

The present study showed the high prevalence of cnLOH in GBMs, indicating that tumorigenesis of GBMs is a complicated process, and that cnLOH appears to play a significant role. Acquired LOH without copy-number abnormality can arise through a number of mechanisms, including monosomic or trisomic rescue (in embryonic development), incomplete segregation of chromosomes, and mitotic nondisjunction.7 The molecular mechanism behind the drastic occurrence of cnLOH in sample 1 is unknown; however, it may result from incomplete segregation of chromosomes by mitotic nondisjunction.7 A recent report on myeloid leukemias also indicated that cnLOH represents a mechanism for oncogenic homozygotic events (activated oncogene or inactivated tumor suppressor gene), without suffering lethal effects from the loss of haploinsufficient genes located within the involved region.8 The same molecular mechanism may be associated with tumorigenesis of these GBM cells in our study.

Our data revealed that more than half of LOH regions in GBMs were cnLOH regions, suggesting that previous studies of CGH analyses may have underestimated the frequency of the genetic perturbation in this disease, because they did not detect cnLOH. Some LOH regions on specific loci, such as combination of 1p/19q losses in anaplastic oligodendrogliomas and 10q losses in GBMs, are well-established prognostic factors.23,28,29 Detection of possible duplication of these chromosomal regions may further improve the prognostic value of genomic diagnosis of this disease.

The use of a high-density SNP array enabled us to evaluate the copy number of neutral LOH regions. However, there are still several challenges, although the application of high-density SNP arrays to simultaneous LOH and copy-number analysis is becoming more popular. One major problem is that homogeneous tumor DNA is necessary to get reliable results, especially in heterogeneous tumors like GBMs, because contamination of nontumor DNA may make accurate LOH and copy-number call in the SNP array difficult. In this study, for example, on chromosome 10 in sample 14 and chromosome 5 in sample 12, copy number is decreased in copy-number analysis while heterozygosity is retained in LOH study (Fig. 2). This phenomenon is possibly due to the presence of contaminating cells in the tumor sample that affected allele calling. In the present study, although we attempted to enrich tumor cells by using laser-capture microdissection, there is still a possibility that extensive macrophages or microglia were present in the dissected tissue because they are morphologically indistinguishable from glioma cells in hematoxylin and eosin staining. The second problem is data analysis. In general, as well as in our study, automated algorithm is usually employed to calculate LOH and copy-number calls from SNP arrays using available software. Although this algorithm provides us an objective standard of analysis, there is a possibility for false-positive or false-negative results because the calling itself is not necessarily correct. We must pay attention to data interpretation particularly in the mosaic pattern of LOH and copy-number results. With these challenges in mind, we strongly believe that our findings of the cnLOH regions in patients would not be the results of such artifacts. Furthermore, our data establish and confirm a genetic contribution of cnLOH to the tumorigenesis or progression of GBMs.

In conclusion, the combination of laser-capture microdissection and high-density SNP array adopted in the present study allowed us to characterize the genomic alterations in GBMs. However, the prognostic consequences of cnLOH for the patients are still unclear. Further studies are required to assess the clinical significance of this phenomenon.

Acknowledgments

This work was supported by a KAKENHI (Grant-in-Aid for Scientific Research) on Priority Areas “Applied Genomics,” KAKENHI (Grant-in-Aid for Exploratory Research) 17659450, and KAKENHI (Grant-in-Aid for Scientific Research) 18390400 from the Ministry of Education, Culture, Sports, Science, and Technology of Japan.

References

1. Stupp R, Mason WP, vanden Bent MJ, et al. Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med. 2005;352:987–996. [PubMed]
2. Konopka G, Bonni A. Signaling pathways regulating gliomagenesis. Curr Mol Med. 2003;3:73–84. [PubMed]
3. Rao RD, James CD. Altered molecular pathways in gliomas: an overview of clinically relevant issues. Semin Oncol. 2004;31:595–604. [PubMed]
4. Hata N, Yoshimoto K, Yokoyama N, et al. Allelic losses of chromosome 10 in glioma tissues detected by quantitative single-strand conformation polymorphism analysis. Clin Chem. 2006;52:370–378. [PubMed]
5. Huang J, Wei W, Zhang J, et al. Whole genome DNA copy number changes identified by high density oligonucleotide arrays. Hum Genomics. 2004;1:287–299. [PMC free article] [PubMed]
6. Zhao X, Li C, Paez JG, et al. An integrated view of copy number and allelic alterations in the cancer genome using single nucleotide polymorphism arrays. Cancer Res. 2004;64:3060–3071. [PubMed]
7. Walker BA, Leone PE, Jenner MW, et al. Integration of global SNP-based mapping and expression arrays reveals key regions, mechanisms, and genes important in the pathogenesis of multiple myeloma. Blood. 2006;108:1733–1743. [PubMed]
8. Fitzgibbon J, Smith LL, Raghavan M, et al. Association between acquired uniparental disomy and homozygous gene mutation in acute myeloid leukemias. Cancer Res. 2005;65:9152–9154. [PubMed]
9. Langdon JA, Lamont JM, Scott DK, et al. Combined genome-wide allelotyping and copy number analysis identify frequent genetic losses without copy number reduction in medulloblastoma. Genes Chromosomes Cancer. 2006;45:47–60. [PubMed]
10. Hu N, Wang C, Hu Y, et al. Genome-wide loss of heterozygosity and copy number alteration in esophageal squamous cell carcinoma using the Affymetrix GeneChip Mapping 10K array. BMC Genomics. 2006;7:299. [PMC free article] [PubMed]
11. Andersen CL, Wiuf C, Kruhøffer M, Korsgaard S, Ørntoft TF. Frequent occurrence of uniparental disomy in colorectal cancer. Carcinogenesis. 2007;28:38–48. [PubMed]
12. Lin M, Wei LJ, Sellers WR, Lieberfarb M, Wong WH, Li C. dChipSNP: significance curve and clustering of SNP-array-based loss-of-heterozygosity data. Bioinformatics. 2004;20:1233–1240. [PubMed]
13. Wang TL, Maierhofer C, Speicher MR, et al. Digital karyotyping. Proc Natl Acad Sci USA. 2002;99:16156–16161. [PMC free article] [PubMed]
14. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-delta delta C(T)) method. Methods. 2001;25:402–408. [PubMed]
15. Yoshimoto K, Iwaki T, Inamura T, Fukui M, Tahira T, Hayashi K. Multiplexed analysis of post-PCR fluorescence-labeled microsatellite alleles and statistical evaluation of their imbalance in brain tumors. Jpn J Cancer Res. 2002;93:284–290. [PubMed]
16. Cianciulli AM, Morace E, Coletta AM, et al. Investigation of genetic alterations associated with development and adverse outcome in patients with astrocytic tumor. J Neurooncol. 2000;48:95–101. [PubMed]
17. Kleihues P, Cavenee WK. World Health Organization classification of tumors: tumors of the nervous system. Lyon, France: IARC Press; 2000. pp. 29–39.
18. Calhoun ES, Hucl T, Gallmeier E, et al. Identifying allelic loss and homozygous deletions in pancreatic cancer without matched normals using high-density single-nucleotide polymorphism arrays. Can Res. 2006;66:7290–7298. [PubMed]
19. Kloth JN, Oosting J, van Wezel T, et al. Combined array-comparative genomic hybridization and single-nucleotide polymorphism-loss if heterozygosity analysis reveals complex genetic alterations in cervical cancer. BMC Genomics. 2007;8:53. [PMC free article] [PubMed]
20. James CD, Carlbom E, Nordenskjold M, Collins VP, Cavenee WK. Mitotic recombination of chromosome 17 in astrocytomas. Proc Natl Acad Sci USA. 1989;86:2858–2862. [PMC free article] [PubMed]
21. Kotliarov Y, Steed ME, Christopher N, et al. High- resolution global genomic survey of 178 gliomas reveals novel regions of copy number alteration and allelic imbalance. Can Res. 2006;66:9428–9436. [PubMed]
22. Wong KK, Tsang YT, Chang YM, et al. Genome-wide allelic imbalance analysis of pediatric gliomas by single nucleotide polymorphic allele array. Can Res. 2006;66:11172–11178. [PubMed]
23. Wiltshire RN, Rasheed BK, Friedman HS, Friedman AH, Bigner SH. Comparative genetic patterns of glioblastoma multiforme: potential diagnostic tool for tumor classification. Neuro-Oncology. 2000;2:164– 173. [PMC free article] [PubMed]
24. Koschny R, Koschny T, Froster UG, Krupp W, Zuber MA. Comparative genomic hybridization in glioma: a meta-analysis of 509 cases. Cancer Genet Cytogenet. 2002;135:147–159. [PubMed]
25. Ohgaki H, Dessen P, Jourde B, et al. Genetic pathways to glioblastoma: a population-based study. Can Res. 2004;64:6892–6899. [PubMed]
26. Rasheed BK, McLendon RE, Friedman HS, et al. Chromosome 10 deletion mapping in human gliomas: a common deletion region in 10q25. Oncogene. 1995;10:2243–2246. [PubMed]
27. Karlbom AE, James CD, Boethius J, et al. Loss of heterozygosity in malignant gliomas involves at least three distinct regions on chromosome 10. Hum Genet. 1993;92:169–174. [PubMed]
28. Ichimura K, Schmidt EE, Miyakawa A, Goike HM, Collins VP. Distinct patterns of deletion on 10p and10q suggest involvement of multiple tumor suppressor genes in the development of astrocytic gliomas of different malignancy grades. Genes Chromosomes Cancer. 1998;22:9–15. [PubMed]
29. Mariani L, Deiana G, Vassella E, et al. Loss of heterozygosity 1p36 and 19q13 is a prognostic factor for overall survival in patients with diffuse WHO grade 2 gliomas treated without chemotherapy. J Clin Oncol. 2006;24:4758–4763. [PubMed]

Articles from Neuro-Oncology are provided here courtesy of Society for Neuro-Oncology and Oxford University Press
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

  • MedGen
    MedGen
    Related information in MedGen
  • PubMed
    PubMed
    PubMed citations for these articles

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...