Cytarabine-Induced Gene Expression Signatures in AML Patients and Its Association with Clinical Outcome.

Abstract 2470

Acute myeloid leukemia (AML) is the second most common form of childhood leukemia and has the one of the worst prognoses of all major childhood cancers. Development of resistance to the nucleoside analog cytarabine (ara-C) is one of the biggest challenges that clinicians face. Studies of diagnostic gene expression profiles have attempted to determine the molecular mechanisms that underlie drug resistance. In the present study, we evaluated cytarabine-induced gene expression changes in n=24 AML patients by using the Affymetrix U133A array. We measured gene expression in bone marrow samples obtained at diagnosis and 24 hours after initiation of cytarabine infusion. We identified genes with significant change in expression post cytarabine treatment (after adjusting for arm) and explored the association of cytarabine induced expression changes with minimal residual disease after one course (MRD), and event-free survival. We found significant change in the expression level of 11 genes at p≤0.001 (6 genes down-regulated and 5 up-regulated) and 146 with p≤0.01 (89 genes down-regulated and 57 up-regulated). Pathway analysis tool (IPA) identified Cancer, Tumor Morphology, Cellular Movement (score 32) and Cell Morphology, Nervous System Development and Function, Tissue Morphology (score 29) as top networks. The most significant pathways influenced by ara-C treatment included NRF2-mediated oxidative stress signaling, Hif 1a signaling, Aryl hydrocarbon receptor and acute phase signaling. The top genes with ∼ 3 fold increase in expression included VAX2, PMP22, GMFG, PTP4A1, TIMP3, CHMP2A, GPR56, NFIB and KLRB1, and top genes with ∼ 3 fold decrease in expression included CER1, PREB, CDK13 and CDK10, UVRAG, FGF17, ZNF768, YLPM1, SIK3, PRICKLE3, MAP3K7.

We also identified 34 genes whose expression change was significantly associated (p<0.001) with clinically beneficial or detrimental pattern of associations using MRD and EFS as clinical endpoints. Additionally 292 genes showed such a pattern of associations at p≤0.01. Analysis of these genes by Ingenuity Pathway identified Cancer, Tissue Development, Connective Tissue Disorders networks (score 43) and DNA Replication, Recombination, and Repair, Nucleic Acid Metabolism, Small Molecule Biochemistry (score 34) as top two networks. Interestingly of the top 34 gene expression changes (p<0.001), all but one gene demonstrated induction in expression post ara-C. Some of the most interesting genes with increase in expression (post ara-C vs. baseline) that were associated with clinical outcome included transcriptional regulators as Runx-related transcription factors RUNX2, and RUNX1T1, which plays role in BMP and TGF-b signaling pathways; NFKB2, which plays role in apoptosis, cell proliferation, differentiation and survival; SMARCA1, which is involved in cell differentiation and remodeling; MKL1 (megakaryoblastic leukemia (translocation) 1), which has been implicated in M7 AML and homeo box proteins HOXC4, HOXD9 and kinases such as.

Overall our results provide insight into cytarabine-induced gene expression differences in AML patients and the association of these expression differences with clinical outcome. Understanding the pathways that are altered in patients once treatment is initiated can provide opportunities to enhance our understanding of drug action and further improve treatment outcome by adjusting the combination therapies.