Characterizing genomic rearrangements in oligodendroglioma using whole genome tilepath hrCGH arrays. N.V. Johnson¹, J.J. Connelly¹, J. Virgadamo¹, R.E. McLendon², J.M. Vance¹, D.D. Bigner², S.G. Gregory¹. 1) Duke Center for Human Genetics, Durham, NC; 2) Duke Comprehensive Cancer Center, Durham, NC.
   Primary brain tumors account for 1% of the new cancer cases in the US which, given the mortality of 4.1% per 100,000 persons, accounted for 13,100 deaths in 2001 alone. These tumors have a broad histopathology, variable sensitivity to treatment and, therefore, have unpredictable progression and survival times. In general, the molecular mechanisms underlying all these variables are poorly understood. For many years it has been contended that genetic instability leads to cancer development via non-random chromosome losses and gains that contribute to tumor malignancy. To identify these underlying molecular mechanisms we have generated high-resolution comparative genomic hybridization (hrCGH) data using our whole genome tilepath microarrays to identify chromosomal rearrangements associated from 110 oligodendroglioma (OD) tumors. We have generated hrCGH data at 100kb resolution from 45 tumors histopathologically determined to be well differentiated oligodendroglioma (WD); 40 anaplastic oligodendroglioma (AO) tumors; and 25 tumors with only an OD designation. Analysis of the entire tumor set identified characteristic loss of 1p and 19q within 72% of the tumors analyzed. In addition to gross chromosome rearrangements, amplifications or deletions (>4Mb) throughout other regions of the genome, we observed a number of single clone deletions and amplifications within our tumor set. The novel data associated with our analysis therefore represents either novel copy number polymorphisms or, more likely, genomic loci which contain genes associated with the development and progression of oligodendroglioma. Additionally, we observed an accumulation of genomic rearrangements between stage II and III stages. We will present a detailed analysis of genomic intervals that define minimally deleted and amplified regions between all OD tumors. These regions now form the focus of on-going candidate gene analysis.