Etiology and Pathogenesis > Molecular OncologyStaging and Prognosis  


40th ASCO Annual Meeting. New Orleans, LA. June 5-8, 2004. Abstract No.8546 (Laboratory Investigation)


Meeting Abstract

Gene expression profiling of progressive and stable childhood pilocytic astrocytomas

L. Girard, J. D. Minna, T. Chen, G. Tomlinson, D. Bowers

University of Texas Southwestern Medical Center, Dallas, TX

Background. Children with Pilocytic Astrocytomas (PAs), the most common CNS tumor in childhood, may have either a quiescent or progressive clinical course. 
At present, extent of tumor resection and tumor location are the most important prognostic factors; biological prognostic factors are unknown. 
The identification of new prognostic factors for PAs may have important clinical and therapeutic implications. 

Methods. Ten mcg of purified mRNA was extracted from archival flash frozen tumor tissue from five completely resected recurrent PAs and six incompletely resected quiescent PAs. 
Samples were hybridized to the Affymetrix GeneChip arrays HG-U133A (22,283 probe sets) and HG-U133B (22,645 probe sets). 
The two arrays together evaluate the expression of 24,698 unique genes. 
Images from scanned chips were processed using Affymetrix software GCOS 1.0 to get the raw signals and p-values for each probe set. 
This data was then analyzed using in-house software (MATRIX): Array data were median-normalized and replicate genes were combined by averaging their signals. 
Samples were then grouped according to tumor recurrence or non-recurrence. 
Signals were averaged within each group and then compared (by calculating log ratios of recurrent vs non-recurrent PAs) and sets of genes showing differential RNA expression in the two groups were derived. 
Unsupervised hierarchical clustering was performed. 

Results. 282 genes were found to be overexpressed at least 2-fold in the subset of progressive PAs compared to the subset of quiescent PAs while 168 genes were found to be underexpressed more than 2-fold. 
Hierarchical clustering did not reveal any association between recurrence and global gene expression. 

Conclusions. These preliminary results demonstrate the ability of gene expression profiling to identify PAs with progressive clinical courses. 
Children with PAs with such profiling may benefit from closer monitoring and/or earlier adjuvant therapy.

Copyright 2004 American Society of Clinical Oncology All rights reserved worldwide.

Source: http://www.asco.org/ac/1,1003,_12-002636-00_18-0026-00_19-003756,00.asp



 

HOME | Detection | Diagnosis | Epidemiology | Etiology & Pathogenesis | Integrative Medicine | Overall Mngt & Case Reports | Prevention | Prognosis | Psychosocial Aspects | Treatment 
About BrainLife
|
Children's Corner | E-mail Alerts | Journals | Newsletter | Patients & Caregivers | Search | Stem Cells | WHO Classification | SITEMAP