This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Pan, E. Articles by Feuerstein, B. G. Search for Related Content PubMed PubMed Citation Articles by Pan, E. Articles by Feuerstein, B. G.

Isochromosome 17q Is a Negative Prognostic Factor in Poor-Risk Childhood Medulloblastoma Patients

Authors' Affiliations: ¹ University of California San Francisco, San Francisco, California; ² Children's Oncology Group, Arcadia, California; ³ Johns Hopkins University School of Medicine, Baltimore, Maryland; and ⁴ Children's Hospital of Philadelphia, Philadelphia, Pennsylvania

Requests for reprints: Shaun Mason, Children's Oncology Group, Publications Office, 440 East Huntington Drive, Suite 300, Arcadia, CA 91066. Phone: 626-447-0064; Fax: 626-445-4334; E-mail: smason{at}childrensoncologygroup.org.

	Abstract

Top Abstract Materials and Methods Results Discussion Conclusions References

 
Background: Medulloblastomas are the most common primary malignant childhood intracranial neoplasms. Patients are currently sorted into three risk groups based on clinical criteria: standard, poor, and infant (<18 months old). We hypothesized that genetic copy number aberrations (CNA) predict prognosis and would provide improved criteria for predicting outcome.

Methods: DNA from 35 medulloblastoma patients from four Children's Cancer Group trials was analyzed by comparative genomic hybridization to determine CNAs. The genetic alterations were evaluated using statistical and cluster analyses.

Results: The most frequent CNAs were gains on 17q, 7, 1q, and 7q and losses on 17p, 10q, X, 16q, and 11q. Amplification at 5p15.1-p15.3 was also detected. Isochromosome 17q (i(17)(q10)) was associated with poor overall survival (P = 0.03) and event-free survival (P = 0.04) independent of poor risk group classification. Age <3 tended to be associated with <3 CNAs (P = 0.06). Unsupervised cluster analysis sorted the study patients into four subgroups based on CNAs. Supervised analysis using the program Significance Analysis of Microarrays (SAM) quantitatively validated those CNAs identified by unsupervised clustering that significantly distinguished among the four subgroups.

Conclusions: Medulloblastomas are genetically heterogeneous and can be categorized into separate genetic subgroups by their CNAs using unsupervised cluster analysis and SAM. i(17)(q10) was a significant independent negative prognostic factor. Infant medulloblastomas may be a distinct genetic subset from those of older patients.

Clinical factors do not predict accurately which standard-risk patients will have early relapses and die. Biological factors, such as elevated TrkC (6) or elevated c-MYC mRNA expression (7) have been correlated with improved and worse survival, respectively. Pomeroy et al. recently reported a multigene predictor based on gene expression profiles that could accurately identify standard-risk medulloblastoma patients with poor clinical outcomes (8). Therefore, we hypothesized that specific genetic alterations are associated with outcome. To test this hypothesis, we analyzed 35 medulloblastoma samples using comparative genomic hybridization (CGH) to detect tumor DNA copy number aberrations (CNA) and correlated the results with patient survival. CGH detects the relative copy number of genetic material by comparing hybridizations of reference and tumor DNA with those of normal metaphase chromosomes (9). Our study goals were to determine whether there were correlations between medulloblastoma CNAs and outcome and use them to develop a more accurate medulloblastoma grading system based on genetic criteria.

	Materials and Methods

Top Abstract Materials and Methods Results Discussion Conclusions References

 
DNA preparation and labeling. Paraffin sections from the Brain Tumor Resource Laboratory of the Children's Cancer Group were deparaffinized by xylene treatment. Standard DNA extraction protocol was applied using a Puregene DNA isolation kit (Gentra Systems, Minneapolis, MN). Normal male and female reference DNA was prepared in the same manner from donor mononuclear cells. Tumor and normal DNA were labeled with fluoroscein-10-dUTP and Texas Red-dUTP (DuPont, Wilmington, DE), respectively, by nick translation using DNase I and Polymerase I from Life Technologies (Carlsbad, CA; ref. 10). Distribution of probe fragment size ranged from 200 to 1,300 bp and the bulk of DNA ranged from 300 to 800 bp. Samples that did not yield enough DNA for nick translation were amplified by degenerate oligonucleotide primer PCR (11). In such cases, the tumor and normal DNA were labeled with digoxygenin-dUTP (Roche, Indianapolis, IN) and FITC-dUTP (DuPont), respectively.

Comparative genomic hybridization. CGH was done according to a previous protocol (10) using 400 ng each of labeled tumor and reference DNA. If degenerate oligonucleotide primer PCR–labeled probes were used for hybridization, the metaphase spreads were incubated first in blocking buffer at room temperature for 5 minutes, then with 1 µg/mL rhodamine-conjugated anti-Dig-antibody (Roche) in blocking buffer at room temperature for 45 minutes. Subsequently, the slides were washed and counterstained with 4',6-diamidino-2-phenylindole. Red, green, and blue images were acquired with a Quantitative Image Processing System, and the ratios of fluorescence intensity along the chromosomes were quantified as described (12).

Statistical methods. We analyzed relationships between clinical variables and CNAs. The primary end points were event-free survival (EFS) and overall survival (OS). EFS was the minimum time from entry to disease progression, second malignant neoplasm, or death from any cause; OS was determined from entry to death from any cause. Nonparametric estimates of EFS and OS probabilities were calculated by the product limit (Kaplan-Meier) estimate, with the SEs computed by the Greenwood formula (13). Estimates are presented as estimate ± SE. Comparisons of outcome were based on the log-rank test and exact log-rank tests from StatXact (Cytel Software Corp, Cambridge, MA). Stratified log-rank tests also were done to adjust for potential prognostic factors. Comparisons of frequency counts were based on two-sided Fisher exact test and ² tests, as appropriate (14). The three patients designated as lost to follow-up did not have continued data forms for follow-up at last contact. All three patients were alive at last contact >5 years from entry date and thus were classified as alive for statistical analysis.

A two-dimensional hierarchical clustering analysis was used to identify groups of patients with similar chromosomal aberrations and common chromosomal aberrations across all cases (15). Uncentered correlation metric and average linkage were used for both dimensions. The relative gains and losses for each chromosome band of all the chromosomes were noted and converted to annotated designations of 0 (no change), +1 (gain), –1 (loss), and +2 (amplification) in chromosomal copy number before clustering.

We used the program Significance Analysis of Microarrays (SAM) to correlate chromosomal aberrations with risk groups, M stage, and gender (16). The percentage of CNAs identified as significant by chance is the False Discovery Rate (FDR). The significance of an aberration is reported as the q value, which is the lowest FDR at which the aberration is designated as significant. We ran SAM using two-class unpaired data analysis with 1,000 permutations. The delta scale was adjusted such that the median number of false significant aberrations was <1.0 and a list of significant CNAs was generated. Those chromosomal bands with qs < 5% were designated as significant.

Case selection. Medulloblastoma cases from the Children's Cancer Group (CCG) protocols were reviewed centrally before inclusion. Initially, 41 cases from five CCG protocols, including protocols accepting patients with tumors other than medulloblastoma, were accepted and analyzed by CGH. Neuropathologists were blinded to original diagnoses and CGH molecular findings. Based on the reviews, three cases were diagnosed as ependymomas, two were atypical rhabdoid tumors, and one had no tumor on the slide and thus could not be histologically verified as medulloblastoma. Genetic changes in these six cases were not characteristic of medulloblastomas. These six cases were excluded based on genetic and histologic analyses, leaving 35 cases in our study. Five of the six excluded cases were from CCG 9921, leaving cases from only four CCG studies (921, 923, 9892, and 9931).

	Results

Top Abstract Materials and Methods Results Discussion Conclusions References

 
Clinical characteristics. The demographic characteristics of the 35 medulloblastoma patients are summarized in Table 1. The median age at study entry was 6.3 years (range, 0.6-14.9 years), there was male predominance, and the distribution of age and M stage was unremarkable. The median number of days on study was 2,696 (range, 48-5,368 days). There was no effect of the various treatment regimens on survival and no significant difference in OS among the four CCG studies (P = 0.78). Five-year OS rates for the four studies were 60 ± 13% for CCG 921, 75 ± 15% for 923, 60 ± 22% for 9892, and 57 ± 19% for 9931. Thus, differences in outcome among the study patients were not likely due to treatment regimens or selection bias. The clinical and genetic characteristics of the medulloblastoma patients and their tumors are listed in Table 2.

View larger version (12K): [in this window] [in a new window]   Fig. 1. Overall survival for all patients.

View larger version (32K): [in this window] [in a new window]   Fig. 2. Summary of gains and losses detected by CGH in 38 cases of paraffin-extracted PNETs.

i(17)(q10) was associated with the poor-risk group (P = 0.01). All five patients with i(17)(q10) were in the poor-risk group and all died. A multivariate Cox regression analysis was used to assess the significance of i(17)(q10) controlled for important prognostic factors, specifically, age (<3 versus 3), M stage (M₀ versus non-M₀), and residual tumor (<1.5 versus >1.5 mL). Based on these factors, increased risk was observed for those with i(17)(q10) with respect to OS (P = 0.004; relative risk, 11.1; 95% confidence interval, 1.9-65.7) as well as EFS (P = 0.011; relative risk, 7.7; 95% confidence interval, 1.4-41.8). In addition, analyses of OS and EFS comparing only the poor-risk patients with and without i(17)(q10) showed significantly worse outcome for the poor-risk patients with i(17)(q10) for both OS (P = 0.024) and EFS (P = 0.034). Five-year OS and EFS rates are 75 ± 15% for the poor-risk patients without i(17)(q10) compared with 20 ± 18% for those with i(17)(q10) (Fig. 3). Thus, even poor-risk patients had reasonable outcomes if their tumors did not have i(17)(q10).

View larger version (11K): [in this window] [in a new window]   Fig. 3. Overall survival by risk groups.

Unsupervised clustering sorted the 35 study patients into four subgroups (A-D; Fig. 4). The 10 tumors in group A tended to have –8, –11p, –11q, –17p, –X, and +7 and +17q. Four of the five i(17)(q10) patients were clustered into group A. Five of the group A patients were standard risk, with one having died. Four of the five poor-risk patients died, and all four had i(17)(q10). Thus, the standard-risk group A patients tended to have more favorable outcomes than the poor-risk group A patients.

View larger version (18K): [in this window] [in a new window]   Fig. 4. Unsupervised cluster analysis.

Group C consisted of 16 tumors with few CNAs. Thirteen of 16 tumors had 3 CNAs. Five had no CNAs, including the three infant patients. There were eight standard-risk patients (seven alive), five poor-risk (four alive), and three infants (all died). The one poor-risk patient who died had +1q, whereas the other poor-risk patients did not gain 1q.

The six tumors in group D are characterized primarily by –9 and –16, and +1q, +6, +7q, and +8. Two of the three standard-risk group D patients died, and two of the three poor-risk patients also died [one of whom had i(17)(q10)]. Both the standard-risk and poor-risk patients who survived had +1q and –9q.

Supervised analysis. Supervised analysis by SAM provides a statistical measure of data associated with references such as gender, M stage, and risk group. The statistical significance is expressed as the q value, which is analogous to the P value, in that the q value decreases with the likelihood that the measurement is actually null. Suppose that features with qs 1.5% are called significant in some genome-wide test of significance. This results in a false-discovery rate of 1.5% among the significant features.

We used clinical criteria such as gender, M₀ versus non-M₀ stage, risk group, overall outcome, and outcome within risk groups to sort patients into groups for SAM analysis. We found that 14q31.1-q31.3 and X were associated with patient gender (q = 2.1, FDR = 2.1). Three female patients had –14q31.1-q31.3, whereas one male patient had gain of this region. Although the female patients had other areas of loss on 14q (Table 2), only 14q31.1-q31.3 was associated with gender. Six female patients had –X, whereas only two male patients had –X. Five of the six female patients with –Xp22.33-q28 survived, whereas the one who died had i(17)(q10). Both male patients died and one had i(17)(q10). All of the patients with CNAs of 14q31.1-q31.3 survived.

Several chromosome regions tended towards associations with EFS. Two patients with +8p12-q24.3 died [one had i(17)(q10)], whereas four patients with –8p12-q24.3 survived (q = 5.1, FDR = 2.6). In addition, six survivors had –Yq11.1-12, whereas none of the patients who died had this loss (q = 5.1, FDR = 2.6).

Unsupervised clustering identified and ranked multiple chromosome regions distinguishing between the subgroups, including 6p, 7p, 7q, 8, 11p, 14q, 15q, 17q, and X. CNAs that best characterized group A by SAM analysis included +7, –8, –11p, +17q, and –X. CNAs characteristic for group B included –14q, –15q, +17q, and –X. There were no CNAs that distinguished group C patients. Group D CNA features included +6p, +7q, and +8. Thus, SAM identified the specific CNAs that determined the medulloblastoma patient genetic subgroups.

SAM did not detect CNAs that distinguished groups by stage, risk group, survival, or age (q > 5) or that distinguished between survivors and nonsurvivors within the standard-risk (q= 81, FDR = 56.5) or poor-risk groups (q = 31, FDR = 21).

	Discussion

Top Abstract Materials and Methods Results Discussion Conclusions References

 
To our knowledge, this was the first study that evaluated the relationship of medulloblastoma survival to CGH. i(17)(q10) was associated with shorter OS (P = 0.03) and EFS (P = 0.04). Five of 13 poor-risk patients had i(17)(q10) (three with disseminated disease and two with postsurgical residual tumors >1.5 mL), all of whom died. Six of the eight poor-risk patients without i(17)(q10) were alive at last contact. Thus, most poor-risk patients had favorable outcomes if their tumors did not have i(17)(q10). A comparison based on only the poor-risk patients showed significant differences in OS (P = 0.024) and EFS (P = 0.034) between patients with and without i(17)(q10). i(17)(q10) has a significant independent negative effect on outcome, and this effect is not driven by the poor-risk classification. Thus, patient outcome, particularly among poor-risk patients, was better determined with genetic criteria compared with clinical criteria alone.

Although the clinical risk group classification is helpful in identifying medulloblastoma patients with worse prognosis, it is known that patients within risk groups have widely varying outcomes. The 5-year OS for poor-risk patients without i(17)(q10) in our study was 75%, which is better than the study group as a whole (63%). Thus, poor-risk classification does not necessitate poor outcome. Furthermore, none of these cases were in the infant risk group, which has a poor prognosis regardless of stage of disease or extent of resection. Because there were relatively few tumors with i(17)(q10), we cannot form a definitive conclusion concerning the role of i(17)(q10) in outcome.

i(17)(q10) is the most common chromosome aberration in childhood medulloblastoma, occurring in 25% to 35% of cases (17–21). Previous analysis suggests it is not an independent prognostic factor (22–24). However, prior studies suggest that poor-risk patients with i(17)(q10) tend toward poor rather than favorable outcomes. For example, Gilbertson et al. described 10 poor-risk medulloblastoma patients with i(17)(q10), six of whom died before study completion (24). Three of the four surviving poor-risk i(17)(q10) patients had desmoplastic medulloblastomas, which reportedly have better outcomes than classic medulloblastomas (20, 25). Only one of the six poor-risk patients who died had a desmoplastic medulloblastoma. In the study by Nicholson et al., four of seven poor-risk i(17)(q10) patients died. One surviving patient in this group had a desmoplastic medulloblastoma, and another had only survived 0.3 years from treatment until the end of the study (20). None of the i(17)(q10) patients who died had desmoplastic medulloblastomas. It is not clear why i(17)(q10) has a negative effect on outcome among poor-risk cases.

Chromosome 17p, the most frequently deleted chromosome (29%) in our study, was not associated with OS (P = 0.14) or EFS (P = 0.15). However, the role of 17p deletion as a prognostic factor remains unclear. Cogen and McDonald reported that deletions on 17p in standard-risk medulloblastoma patients were associated with poor outcome (P = 0.022; ref. 26), whereas Biegel et al. found that 17p deletion tended to associate poor outcome but the trend was not statistically significant (27). Emadian et al. found no association between 17p deletion and survival (28). Eberhart et al. reported that all seven cases with 17p loss in their study occurred in anaplastic medulloblastomas, four of whom died within 27 months from diagnosis (29). Thus, 17p loss may be associated with more biologically aggressive medulloblastomas.

The second most common aberration in medulloblastomas is gain of the entire chromosome 7. Although the frequency of i(17)(q10) in this study (14%) was lower than previously, the frequency of +7 (23%) was comparable with the reported rate of 25% (18–21, 30–33). +7 was not associated with OS (P = 0.26) or EFS (P = 0.25) in our study. +7 is a frequent aberration in glioblastomas (34) and is associated with worse survival in anaplastic astrocytoma patients (35), but its significance in medulloblastomas is unknown.

Deletion of 9q tended to associate with improved OS and EFS (P = 0.07). Four of the five patients were standard risk and all five were alive at last contact. This observation may be related to observations associating mutations of the PTCH gene (9q22.3) with desmoplastic medulloblastomas (36, 37). Pediatric patients with desmoplastic medulloblastomas may have more favorable outcomes than classic medulloblastoma patients (20, 25). There were four patients in our study (two standard and two poor risk) who had non-nodular desmoplastic medulloblastomas; all four were alive at last contact.

CNA number tended to associate with age <3 (P = 0.06). Four patients were <3 years of age, three of whom were infants. None of these three tumors had CNAs, and none survived longer than 285 days from study entry. The fourth was a poor-risk patient who had two CNAs and survived. Medulloblastomas in infants may represent a genetic subset distinct from medulloblastomas in older children.

One tumor had amplification of 5p15.1-p15.3. Amplification at 5p15 was reported in patients with good outcomes (18, 33). Amplification of the hTERT gene (chromosome 5p15.33), which encodes the catalytic protein subunit of telomerase and is associated with carcinogenesis, has been reported in 38% of medulloblastomas (38). In our study, the patient with 5p amplification was still alive over 12 years after diagnosis.

Unsupervised clustering sorted our study patients into four subgroups. Several CNAs were identified as defining the various subgroups. Group C patients were characterized by few CNAs. Because unsupervised clustering is a qualitative rather than quantitative analysis, statistical correlations could not be made with regard to associations between particular CNAs and their subgroups or CNAs to outcome.

SAM analysis quantitatively validated the results of our unsupervised cluster analysis. CNAs that significantly distinguished between the subgroups were also identified by unsupervised clustering. There were also CNAs that distinguished between gender (female patients tended to have –14q31.1-q31.3 and –X). Given the small number of patients, definitive conclusions should not be made regarding the significance of these particular CNAs. However, the application of SAM to further define genetic medulloblastoma subgroups is appropriate and useful for future studies.

	Conclusions

Top Abstract Materials and Methods Results Discussion Conclusions References

 
Our study had the largest reported number of medulloblastoma cases with CGH and outcome data to our knowledge. i(17)(q10) is associated with poor outcome independent of risk group classification, and low number of CNAs tends to be associated with poor outcome. Medulloblastomas in infants may be a genetically distinct subset from those in older children. Our results showed that medulloblastomas are genetically heterogeneous tumors. Genetic aberrations define subsets of medulloblastomas beyond clinical criteria and seem to have a stronger predictive value for outcome compared with clinical criteria alone, particularly for poor-risk patients. In our study, poor-risk patients without i(17)(q10) had significantly improved OS and EFS compared with poor-risk i(17)(q10) patients. As noted in Case Selection, genetic analysis can help identify histologically misdiagnosed tumors. There was a strong concordance in our study between the results of unsupervised cluster analysis and supervised analysis with SAM. These analyses suggest specific chromosomes are important for determining medulloblastoma patient subsets, but the data must be interpreted cautiously until larger studies verify these results. Subsequent studies should include a larger number of patients analyzed with array CGH, which has significantly higher resolution than standard CGH, and should employ not only standard statistical methods but also unsupervised cluster analysis and SAM. Future studies that accrue sufficient patient numbers may confirm specific CNAs as prognostic factors for childhood medulloblastoma patients. These results combined with gene expression analyses and current clinical stratification criteria may lead to the development of a highly accurate predictive model for medulloblastoma behavior and outcome.

	Acknowledgments

 
We thank Luanne Wainwright for her technical assistance; Richard Sposto, Ph.D. for his assistance with statistical analysis; Steve Qualman, MD for his assistance with the tumor slide review; Shaun Mason for article review; and the Cooperative Human Tissue Network.

	Footnotes

 
Grant support: NIH grants T32 CA09291, CCG U10 CA13539, NS42927, NBTF CA85799, and NIH CA 46274.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received 3/ 3/04; revised 3/18/05; accepted 4/ 7/05.

	References

Top Abstract Materials and Methods Results Discussion Conclusions References

 

1. Rutka JT. Medulloblastoma. Clin Neurosurg 1997;44:571–85.[Medline]
2. Packer RJ, Sutton LN, Elterman R, et al. Outcome for children with medulloblastoma treated with radiation and cisplatin, CCNU, and vincristine chemotherapy. J Neurosurg 1994;81:690–8.[Medline]
3. Eberhart CG, Kepner JL, Goldthwaite PT, et al. Histopathologic grading of medulloblastomas: a Pediatric Oncology Group study. Cancer 2002;94:552–60.[CrossRef][Medline]
4. McManamy CS, Lamont JM, Taylor RE, et al. Morphophenotypic variation predicts clinical behavior in childhood non-desmoplastic medulloblastomas. J Neuropathol Exp Neurol 2003;62:627–32.[Medline]
5. Geyer JR, Zeltzer PM, Boyett JM, et al. Survival of infants with primitive neuroectodermal tumors or malignant ependymomas of the CNS treated with eight drugs in 1 day: a report from the Children's Cancer Group. J Clin Oncol 1994;12:1607–15.[Abstract/Free Full Text]
6. Kim JY, Sutton ME, Lu DJ, et al. Activation of neurotrophin-3 receptor TrkC induces apoptosis in medulloblastomas. Cancer Res 1999;59:711–9.[Abstract/Free Full Text]
7. Grotzer MA, Hogarty MD, Janss AJ, et al. MYC messenger RNA expression predicts survival outcome in childhood primitive neuroectodermal tumor/medulloblastoma. Clin Cancer Res 2001;7:2425–33.[Abstract/Free Full Text]
8. Pomeroy SL, Tamayo P, Gaasenbeek M, et al. Prediction of central nervous system embryonal tumour outcome based on gene expression. Nature 2002;415:436–42.[CrossRef][Medline]
9. Kallioniemi A, Kallioniemi OP, Sudar D, et al. Comparative genomic hybridization for molecular cytogenetic analysis of solid tumors. Science 1992;258:818–21.[Abstract/Free Full Text]
10. Mohapatra G, Kim DH, Feuerstein BG. Detection of multiple gains and losses of genetic material in ten glioma cell lines by comparative genomic hybridization. Genes Chromosomes Cancer 1995;13:86–93.[Medline]
11. Hirose Y, Aldape K, Takahashi M, et al. Tissue microdissection and degenerate oligonucleotide primed-polymerase chain reaction (DOP-PCR) is an effective method to analyze genetic aberrations in invasive tumors. J Mol Diagn 2001;3:62–7.[Abstract/Free Full Text]
12. Piper J, Rutovitz D, Sudar D, et al. Computer image analysis of comparative genomic hybridization. Cytometry 1995;19:10–26.[CrossRef][Medline]
13. Kalbfleisch J, Prentice R. The statistical analysis of failure time data. New York: John Wiley and Sons; 1980.
14. Dixon W, Massey F. Introduction to statistical analysis. 3rd ed. New York: McGraw-Hill; 1969.
15. Eisen MB, Spellman PT, Brown PO, et al. Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci U S A 1998;95:14863–8.[Abstract/Free Full Text]
16. Tusher VG, Tibshirani R, Chu G. Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci U S A 2001;98:5116–21.[Abstract/Free Full Text]
17. Biegel JA, Rorke LB, Packer RJ, et al. Isochromosome 17q in primitive neuroectodermal tumors of the central nervous system. Genes Chromosomes Cancer 1989;1:139–47.[Medline]
18. Reardon DA, Michalkiewicz E, Boyett JM, et al. Extensive genomic abnormalities in childhood medulloblastoma by comparative genomic hybridization. Cancer Res 1997;57:4042–7.[Abstract/Free Full Text]
19. Avet-Loiseau H, Venuat AM, Terrier-Lacombe MJ, et al. Comparative genomic hybridization detects many recurrent imbalances in central nervous system primitive neuroectodermal tumours in children. Br J Cancer 1999;79:1843–7.[CrossRef][Medline]
20. Nicholson JC, Ross FM, Kohler JA, et al. Comparative genomic hybridization and histological variation in primitive neuroectodermal tumours. Br J Cancer 1999;80:1322–31.[CrossRef][Medline]
21. Russo C, Pellarin M, Tingby O, et al. Comparative genomic hybridization in patients with supratentorial and infratentorial primitive neuroectodermal tumors. Cancer 1999;86:331–9.[CrossRef][Medline]
22. Nam DH, Wang KC, Kim YM, et al. The effect of isochromosome 17q presence, proliferative and apoptotic indices, expression of c-erbB-2, bcl-2 and p53 proteins on the prognosis of medulloblastoma. J Korean Med Sci 2000;15:452–6.[Medline]
23. DeChiara C, Borghese A, Fiorillo A, et al. Cytogenetic evaluation of isochromosome 17q in posterior fossa tumors of children and correlation with clinical outcome in medulloblastoma. Detection of a novel chromosomal abnormality. Childs Nerv Syst 2002;18:380–4.[CrossRef][Medline]
24. Gilbertson R, Wickramasinghe C, Hernan R, et al. Clinical and molecular stratification of disease risk in medulloblastoma. Br J Cancer 2001;85:705–12.[CrossRef][Medline]
25. Sure U, Berghorn WJ, Bertalanffy H, et al. Staging, scoring and grading of medulloblastoma. A postoperative prognosis predicting system based on the cases of a single institute. Acta Neurochir 1995;132:59–65.
26. Cogen PH, McDonald JD. Tumor suppressor genes and medulloblastoma. J Neurooncol 1996;29:103–12.[CrossRef][Medline]
27. Biegel JA, Janss AJ, Raffel C, et al. Prognostic significance of chromosome 17p deletions in childhood primitive neuroectodermal tumors (medulloblastomas) of the central nervous system. Clin Cancer Res 1997;3:473–8.[Abstract]
28. Emadian SM, McDonald JD, Gerken SC, et al. Correlation of chromosome 17p loss with clinical outcome in medulloblastoma. Clin Cancer Res 1996;2:1559–64.[Abstract]
29. Eberhart CG, Kratz JE, Schuster A, et al. Comparative genomic hybridization detects an increased number of chromosomal alterations in large cell/anaplastic medulloblastomas. Brain Pathol 2002;12:36–44.[Medline]
30. Schutz BR, Scheurlen W, Krauss J, et al. Mapping of chromosomal gains and losses in primitive neuroectodermal tumors by comparative genomic hybridization. Genes Chromosomes Cancer 1996;16:196–203.[CrossRef][Medline]
31. Bayani J, Zielenska M, Marrano P, et al. Molecular cytogenetic analysis of medulloblastomas and supratentorial primitive neuroectodermal tumors by using conventional banding, comparative genomic hybridization, and spectral karyotyping. J Neurosurg 2000;93:437–48.[Medline]
32. Gilhuis HJ, Anderl KL, Boerman RH, et al. Comparative genomic hybridization of medulloblastomas and clinical relevance: eleven new cases and a review of the literature. Clin Neurol Neurosurg 2000;102:203–9.[CrossRef][Medline]
33. Michiels EM, Weiss MM, Hoovers JM, et al. Genetic alterations in childhood medulloblastoma analyzed by comparative genomic hybridization. J Pediatr Hematol Oncol 2002;24:205–10.[CrossRef][Medline]
34. Liu L, Ichimura K, Pettersson EH, et al. Chromosome 7 rearrangements in glioblastomas; loci adjacent to EGFR are independently amplified. J Neuropathol Exp Neurol 1998;57:1138–45.[Medline]
35. Kunwar S, Mohapatra G, Bollen AW, et al. Genetic subgroups of anaplastic astrocytomas correlate with patient age and survival. Cancer Res 2001;61:7683–8.[Abstract/Free Full Text]
36. Raffel C, Jenkins RB, Frederick L, et al. Sporadic medulloblastomas contain PTCH mutations. Cancer Res 1997;57:842–5.[Abstract/Free Full Text]
37. Pietsch T, Waha A, Koch A, et al. Medulloblastomas of the desmoplastic variant carry mutations of the human homologue of Drosophila patched. Cancer Res 1997;57:2085–8.[Abstract/Free Full Text]
38. Fan X, Wang Y, Kratz J, et al. hTERT gene amplification and increased mRNA expression in central nervous system embryonal tumors. Am J Pathol 2003;162:1763–9.[Abstract/Free Full Text]

This article has been cited by other articles:

				K. C. Lo, C. Ma, B. N. Bundy, S. L. Pomeroy, C. G. Eberhart, and J. K. Cowell Gain of 1q Is a Potential Univariate Negative Prognostic Marker for Survival in Medulloblastoma Clin. Cancer Res., December 1, 2007; 13(23): 7022 - 7028. [Abstract] [Full Text] [PDF]

				F. Mendrzyk, B. Radlwimmer, S. Joos, F. Kokocinski, A. Benner, D. E. Stange, K. Neben, H. Fiegler, N. P. Carter, G. Reifenberger, et al. Genomic and Protein Expression Profiling Identifies CDK6 As Novel Independent Prognostic Marker in Medulloblastoma J. Clin. Oncol., December 1, 2005; 23(34): 8853 - 8862. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Pan, E. Articles by Feuerstein, B. G. Search for Related Content PubMed PubMed Citation Articles by Pan, E. Articles by Feuerstein, B. G.