Abstract 3209: Gene expression micro array analysis of diagnosis and matched relapse pediatric AML samples indicates that immune regulatory pathways and epigenetic factors are involved in disease progression

Abstract

Abstract Due to optimization of induction treatment, approximately 90% of AML pediatric patients achieve a complete remission. However, relapses still occur in 30-40% of patients, with dismal outcome. In order to find new options for targeted treatment, we aim to identify key molecules and signaling pathways involved in the development of relapses. To this end we use gene expression profiling. Total mRNA of initial diagnosis and matched relapse bone marrow samples (>80% blasts) was obtained from 27 pediatric AML patients and used for Affymetrix HGU 133 plus 2.0 microarrays. For statistical analyses we used BioConductor packages, Significance Analysis of Microarrays (SAM) and BRB Arraytools. Pathway analysis was performed using Ingenuity and Gene Set Enrichment Analysis (GSEA). Analysis of paired diagnosis and relapse samples revealed large inter-patient differences in the number of genes that were differentially expressed. Unsupervised cluster analysis showed that for 47% of patients the diagnosis and relapse sample cluster next to each other. The remaining paired samples clustered more distant and interestingly the majority of these pairs also showed changes in mutation status of FLT3 or RAS. The former have been described by us to be associated with shorter time to relapse (Bachas et al, Blood, 2010). Pathway analysis of the differentially expressed genes (> 1.5 fold) revealed molecular pathways implicated in cancer, inflammatory disease, hematopoietic development and genetic disorders. Involved biological processes were related to the immune system, nuclear organization and intracellular trafficking. We performed in silico class prediction and found 31 genes to be differentially expressed and discriminative for diagnosis or relapse samples with an accuracy of more than 80%, as determined by ‘leave-one-out’ cross validation. Of the 31 genes, 29 showed a lowered expression in the relapse sample when compared to the initial diagnosis sample (median intensity ratio diagnosis/relapse = 1.74, range=1.47-2.33, P29th gene=0.032) and 7 of these 29 genes have functions in the maintenance of chromatin structure. Preliminary RT-PCR results confirmed this for an independent set of patient samples. In vitro cytotoxicity experiments on primary patient material using relevant experimental drugs are underway. In conclusion, we identified genes and pathways that were significantly differentially expressed between diagnosis and relapse samples. The majority of discrimative genes are down-regulated at relapse with an important contribution of genes involved in maintenance of chromatin structure. Our efforts are directed to determine if these findings will be instrumental to design therapies aimed at preventing relapses. Financially supported by the Dutch Cancer Society (VU 2005-3666, J.C.) Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 3209. doi:10.1158/1538-7445.AM2011-3209