Integrative analysis of gene expression profiling and high-resolution genomic copy number in gastrointestinal stromal tumor (GIST)

10559 Background: Besides mutually exclusive cKIT or PDGFRA mutations, sequential accumulation of other genetic events may be involved in GISTs development and progression, but very few data is still available. Methods: Fresh tissues specimens of GISTs from 10 patients (9 gastric and 1 intestinal) were collected and used for RNA and DNA extraction, labeled and hybridized to HG-U133Plus 2.0 and SNP array 6.0, respectively (Affymetrix). Six patients had exon 9 or exon 11 c-KIT mutation, two PDGFRA mutation, and other two wild-type disease. Gene expression data were quantified by the RMA algorithm, filtered and analysed with supervised techniques (SAM algorithm). Genomic copy number data were analysed with Partek Genomic Suite software against a reference set of 90 Ceu HapMap individuals using a segmentation algorithm with stringent p-value cutoff. Results: Almost all patients exhibited both macroscopic cytogenetic alterations and cryptic microdeletions or amplifications by SNP-array copy number data analysis. The most frequent chromosomal alterations were: 14q complete or partial deletion (7/10), chromosome 19 monosomy (3/10), 22q and 1p deletion (2/10), chromosome 5 trisomy (2/10). The minimal overlapping region ranged from 14q22.3 to 14q32.33, covering a region including 320 genes. The integration of copy number and gene expression data showed that at least 40% of the genes inside the 14q deleted region were significantly downregulated (FDR<10%) in comparison to 14q-diploid patients. In this region several tumor suppressor genes involved in cell cycle checkpoint control (SNW1, CHES1, PPP2R5E), apoptosis induction (PPM1A, MOAP1, PPP1R13B), DNA damage response (MLH3, TDP1), WNT/Notch pathway inhibition (NUMB, DACT1, SEL1L) are located. Conclusions: A wide spectrum of genetic aberrations in GISTs may occur besides c-KIT and PDGFRA mutations. The most frequent is 14q deletion that leads to a significant downregulation of many putative tumor suppressor genes. Combining gene expression and high resolution genomic copy number analysis could identify new haploinsufficient tumor suppressor genes involved in GISTs pathogenesis and tumor progression. No significant financial relationships to disclose.