Detection of submicroscopic chromosomal aberrations and allelic alterations in unpaired tumor samples by genome-wide SNP scan

Chromosomal aberrations and allelic alterations in cancer 
genomes are important hints lead to discovery of oncogenes and 
tumor suppressor genes. However, normal cell contamination and 
clonal heterogeneity in tumor specimen complicate the detection 
of these events. Besides, short of paired normal tissue limits
the application of loss of heterozygousity (LOH) detection. We 
have developed a paradigm which combined copy number estimation 
and a new algorithm for allelic imbalance for whole-genome SNP
genotyping data analysis to reveal submicroscopic chromosomal
aberrations and allelic alterations in unpaired tumor samples.
We have applied this paradigm to several types of cancers. I will
illustrate the concept and results using acute lymphoblastic
leukemia as an example. 
	A total of 12 T-ALL patients, 2 B-ALL patients, and one pair
of human primary tumor and mouse engraft from the same T-ALL
patient have been studied. A total of 54 non-redundant
cancer-specific regions, including 26 ≤ 4Mb, were identified;
six of the 26 submicroscopic regions were novel as they have
never been reported in T-ALL, including amplification of the
MYCN locus and deletion of the RUNX1 (AML1) locus. Furthermore,
allelic imbalance analysis allowed the identification of 
complex deletions and amplifications that might reflect clonal 
heterogeneity, and detection of copy-neutral loss of 
heterozygosity in the unpaired tumor samples, with the involved 
alleles assigned. Moreover, the genomic contents from the pair
of human primary tumor and mouse engraft were very similar.
Interestingly, enhanced signals of several chromosomal
aberrations in mouse engraft were observed, suggesting clonal 
selection during cancer growth. The chromosomal aberrations
were verified using real-time PCR and fluorescence in-situ
hybridization.
	These multi-dimension results from SNP genotyping will
facilitate the identification of candidate genes involved in 
leukaemogenesis. Moreover, the dissection of tumor
heterogeneity and the characterization of mouse engraft at the
molecular level will provide valuable insight into the complex
process of tumor evolution and therapeutic targeting.