Clinical responses in epigenetic therapy with MS275 and azacytidine are associated with activation of signaling pathway genes

2481

Treatment with DNA methyltransferase inhibitors and histone deacetylase inhibitors leads to re-expression of genes transcriptionally silenced by promoter methylation. A phase I study combining 5-azacytidine and MS-275 was conducted in patients with myelodysplastic syndrome and acute myeloid leukemia. Subcutaneous 5-azacytidine (30-50 mg/m2/dose) was administered for 10 days. MS-275 (2 -8 mg/m2/dose) was administered orally days 3 and 10 of each 28 day treatment cycle. RNA was extracted from CD34+ cells from 18 patients (9 responders, 9 non-responders) pre-treatment and at day 15 and 29 day after the first cycle and hybridized to Affymetrix U133V2 chips. Data were normalized by GC-RMA and experimental batch effect was adjusted by a Bayesian method. Differentially expressed genes were identified by the Students t-test using a cutoff of p≤0.01. Biological functions and gene interaction networks denoted by differential expressed genes were analyzed using, Ingenuity Pathways Analysis and GeneSet Enrichment Analysis (GSEA). Unsupervised clustering of gene expression patterns at day 0, 15 or 25 did not distinguish responders from non-responders. Pair-wise comparisons of gene expression showed that 238 genes were significantly regulated at day 15, while 122 genes were found in at day 29 when compared to day 15. There was almost no overlap between these sets, with only 2 genes significantly regulated at both timepoints. Gene Ontology analysis indicated that the genes involved in immune response and cell signaling were over-presented at day 15, while genes involved in cellular metabolism were induced at day 29, indicating that the drugs may affect different mechanisms over time. None of a set of genes assayed for demethylation showed evidence of gene methylation changes or expression changes after therapy. Supervised gene expression analysis by clinical response showed that 67 genes were differentially expressed at baseline between resp vs non-resp, after treatment this increased this to 114 and 169 genes on days 15 and 29. Only 2 genes different at baseline remained differentially expressed in the two groups. GSEA and GO analysis showed that responders had low basal levels of genes involved in tyrosine kinase signaling and these were up-regulated by day 15 post-treatment. Furthermore genes involved NFκB activity were significantly lower in the responders than the non-responders at 15th day after treatment. Given the frequency of anomalies of tyrosine kinase pathways in MDS and AML, this suggests that subsets of patients with active basal levels of signaling and who maintain NFκB activity after therapy may be less amenable to epigenetic therapy. These data suggest that therapies targeting to signaling pathways or the NFκB axis might be considered along with epigenetic therapy. Ongoing integrative analyses will combine global gene methylation analysis with gene expression analysis.