Abstract 600

Background:

Azacitidine is hypothesized to exert its therapeutic effect in patients with myelodysplastic syndrome (MDS) through inhibition of DNA methylation. However to date no genomic DNA methylation pattern has been shown to predict response to azacitidine in patients with MDS, and no aberrantly silenced gene or group of genes has been shown to be reactivated by azacitidine that can be clearly linked to the beneficial clinical effect. We sought to identify the gene or group of aberrantly hypermethylated genes that are responsible for the therapeutic effect of azacitidine by retrospectively analyzing genome-wide DNA methylation profiles from bone marrow samples of a cohort of 113 patients with MDS treated with the DNA methylation inhibitor, azacitidine.

Methods:

Bone marrow aspirates were collected at time of diagnosis prior to treatment, after 4 cycles of azacitidine therapy and 8 cycles of therapy. DNA was isolated and bisulfite treated with the EZ-96 DNA Methylation-Gold Kit. DNA methylation analysis was performed on 27,578 CpG sites representing 14,475 genes (almost ¾ of known genes) using the Infinium Bead Array system for samples at the time of diagnosis, 4 and 8 cycles of therapy. Only 19,662 CpG sites were used for further analysis due to exclusion of CpG sites that were on the × chromosome, sites suspected of containing single nucleotide polymorphisms (SNP), and sites within DNA repeats. In total 91 samples were analyzed from 43 patients with MDS, which were selected to represent different disease classifications and responses to therapy, and bone marrow aspirates from 10 healthy control subjects without MDS.

Results:

Two-way hierarchical cluster analysis showed clear clustering of bone marrow samples taken from subjects without MDS. DNA methylation patterns from healthy controls clustered together, and pre and post azacitidine treatment samples from the same subject clustered together as well. Samples did not cluster by DNA methylation patterns for WHO classification, International Prognostic Scoring System (IPSS), cytogenetic abnormalities, or response to azacitidine. Supervised cluster analysis is ongoing. Global decreases in DNA methylation as measured by the average methylation for all 19,662 loci assayed did decrease with treatment and there was a trend for a larger decrease in DNA methylation in those patients who responded to azacitidine.

Conclusion:

In this pilot study of genome-wide DNA methylation analysis of MDS patients treated with azacitidine we find global decreases of DNA methylation. We were unable to identify a DNA methylation pattern or group of hypermethylated genes that would predict response to azacitidine. MDS samples did not cluster by WHO classification, IPSS or response to azacitidine. Larger translational studies are needed, but the possibility that DNA methylation decreases in patients treated with azacitidine serve as a pharmacological marker rather than a therapeutic target should also be considered

Disclosures:

Laird:Celgene: Consultancy. Yang:Celgene: Honoraria, Research Funding, Speakers Bureau.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution