Funding Agency:

Children's Brain Tumor Foundation

Total Project Period:

11/1/07 - 11/30/09

Total Project Award:

$150,000

Principal Investigator:

Ian F. Pollack, MD

Co-Investigators:

Ronald Hamilton, MD

Project Summary:

The specific aims of this project are:

1. To evaluate the prognostic significance of a hypothesis-based panel of molecular and biological markers in pediatric oligodendrogliomas, and
2. To correlate outcome with clinical correlates, such as tumor location, and resection extent.

Low-grade gliomas are the most common brain tumors in children. Affected patients have a widely variable prognosis.  While most children with superficial cerebral lesions are cured with gross total resection, lesions in deep or functionally critical sites often cannot be completely removed and carry high risks for recurrence and morbidity. To date, there has been a paucity of data regarding molecular or biological factors associated with outcome and insights regarding therapeutic targets have been lacking. Pediatric oligodendrogliomas in particular have been poorly characterized on a molecular and biological basis, and it remains uncertain whether markers with prognostic significance in adults also have predictive value in children or, conversely, whether other markers are more important. This determination has been hampered by the fact that these tumors occur less frequently than in adults and that pediatric molecular analyses to date have focused on other more common lesions, such as medulloblastoma, ependymoma, and malignant glioma.

To address this limitation, the proposed study would take advantage of a sizeable cohort of pediatric oligodendrogliomas at our institution for which biological material at diagnosis is available. This group includes more than 70 cases with matched histological material and outcome data extending in some cases for more than 20 years. Although a number of these cases have been analyzed ad hoc for various biological correlates, there have never been available funds to examine this unique resource in a systematic fashion, using the broad panel of genotyping and phenotyping technology that we have extensively applied for pediatric high-grade gliomas, as noted in our preliminary studies. The studies included in the proposed study are designed to address this issue. This analysis should have sufficient statistical power to identify meaningful prognostic associations, and would provide new insights into biological correlates of prognosis in pediatric oligodendrogliomas that may improve the chances for curing these tumors.

Preliminary Studies

We have had extensive experience involving molecular marker analysis in pediatric high-grade gliomas. This work began with initial multiparametric screening of a wide range of potential markers in an institutional cohort of childhood high-grade gliomas. These studies included immunohistochemical analysis of the expression of p53, basic fibroblast growth factor, EGFR, bcl-2, p16, MIB-1, and MGMT, as well as microdissection-based genotyping of a large number of targets, searching for point mutational change, allelic imbalance, and/or altered gene copy number, depending upon the gene under study. Our analysis incorporated genes that have been implicated in glial neoplasia, and a variety of other markers that have not been previously examined in pediatric gliomas. For targets of particular interest, sequencing analysis was coupled with loss of heterozygosity analysis or fluorescence in situ hybridication.  Our analysis in this cohort demonstrated a strong association between p53 overexpression and mutations and outcome, and between proliferation labeling and outcome. 

We extended this analysis to the multi-institutional CCG cohort of 945, the largest study for childhood high-grade gliomas ever completed, and then applied these analyses prospectively to the ACNS0126 and ACNS0423 cohorts.  These studies confirmed an association between p53 status and outcome, demonstrated an association between p16 deletions and survival, highlighted a correlation between proliferation index and outcome, and demonstrated significant molecular distinctions between pediatric and adult high-grade gliomas. This analysis also showed a striking association between expression of methylguanine-DNA methyltransferase (MGMT) and outcome in the CCG-945 cohort, which was confirmed in the ACNS0126 cohort. We have extended our analysis to high-throughput SNP micro-array-based allelotyping, using whole genome amplification to allow analysis of archival, paraffin-embedded material using the Affymetrix platform enab.