1699P - Identifying Trans-Acting Copy-Number Alterations in Lung Adenocarcinomas

Abstract

ABSTRACT Background One of the main challenges in genomic oncology is to identify somatic copy-number alterations (CNAs) that include critical genes driving either the initiation or the progression of cancer disease. Most integrated genomic/transcriptomic analyses have primarily focused on identifying cis-acting CNAs, whose copy gains or losses are significantly associated with gene expression changes for the genes they regulate. In this work, we do not restrict our analysis to cis-acting CNAs but focus on CNAs whose changes are associated with large number of significant transcriptomic changes distributed over the entire genome. We hypothesize that CNAs influencing a large number of transcripts either by direct or indirect effect are likely to harbor candidate targetable genes. Material and methods A homogeneous series of 129 early stage lung adenocarcinomas, profiled using both high-resolution array-based genotyping (SNP-array) and gene expression, was analyzed. After classical preprocessing, we selected recurrent CNAs that were exclusively amplified or deleted using a previously published latent class model analysis (BMC Med Genomics. 2009 Jul 14;2:43). For each genomic area, we computed genome-wide statistics measuring the relationship between transcriptomic changes over the different levels of copy number changes (copy loss/modal/copy gain). Then, we reported the number of significant associations taking into account multiple testing. We finally focused on “trans-acting CNAs” defined as those having extreme values. Results Main trans-acting CNAs were found on 1q, 4q, 5q, 9p, 14q, harboring several known oncogenes/tumor suppressor genes. In contrast, other classical recurrent CNAs (such as 5p amplification) were not selected, suggesting they were not associated with significant phenotype changes, as defined by transcriptomic analysis. The relationship between each “trans-acting CNAs” and clinico-pathological variables, as well as outcome is discussed. Conclusion Using an integrative high-throughput microarray analysis in a series of lung adenocarcinomas, we identify “trans-acting CNAs”. We show the interest of focusing on these CNAs which are likely to harbor potentially targetable candidate genes. Disclosure All authors have declared no conflicts of interest.