2002 CTOS Annual Meeting Posters — Biology

CDNA MICROARRAY ANALYSIS OF MESENCHYMAL NEOPLASIA EX VIVO
[Abstract ID: 16]

Category: Biology

Authors: R. Lor Randall¹, Mark Wade¹, Karen Albritton¹

Author Institutions: ¹Huntsman Cancer Institue at the University of Utah, Utah, United States

Presenter: R. Lor Randall
r.lor.randall@hsc.utah.edu

Correspondent: R. Lor Randall
r.lor.randall@hsc.utah.edu
SLC Utah United States 84112-5550
Ph: 801-585-0300
Fax: 801-585-0159

Objectives: Mesenchymal neoplasms can be a diagnostic dilemma in 17-48% of cases. By characterizing and scrutinizing genetic profiles via cDNA microarray technology, we will better understand the relationship and etiology of these tumors, facilitating diagnosis and therapeutics. We present our experience to date with emphasis on technical details in performing cDNA microarray analysis on samples from actual solid tumors ex vivo. Goals: 1) validate serial arrays from a given heterogeneous tumor are reproducible across preparations; 2) verify tumor specific heterogeneity is not significant relative to across a) tumor types b) individuals. Establish standards to identify tumor specific deviant genes using hierarchial clustering and principal component analysis.

Methods: Tissue is acquired using IRB approved methods from untreated primary neoplasms and immediately placed in stabilizing solution. Total RNA is extracted with mRNA isolation carried out within 72 hours. Samples are labeled and hybridized to microarrays containing 6912 probes. Three cohorts (Desmoid, MPNST and GCT) are presented. Cohorts were analyzed using hierarchial clustering (HC) and principle component analysis (PCA).

Results: Of seventy-four prospective tissues inventoried, thirty-six (35%) yielded mRNA greater the 230 micrograms. The average amount of mRNA extracted was 17.1 micrograms/gram of tissue. (S.D.= 9.9 ug/g; range 1.1-38.8ug/g). HC and PCA demonstrated reproducible clustering within a given heterogenous tumor type and compared to across tumor types and individuals.

Conclusions: Serial arrays from a given heterogeneous tumor are reproducible across preparations. Tumor specific heterogeneity is not significant relative to across tumor types and individuals. This technology is a viable tool to improve diagnostic discrepency in mesenchymal neoplasia.