Azacitidine (AZA) has been shown to improve overall survival in patients with myelodysplastic syndrome (

Silverman LR et al.
J Clin Oncol
2002
;
20
:
2429
–40
). AZA is postulated to work, in part, by inducing hypomethylation and thereby altering gene expression. As a correlative analysis to our phase II study of IV AZA we assessed the baseline methylome of MDS and the changes in global methylation on cycle 2 day 25 (c2d25) of treatment with AZA (75mg/m2 d1–5 of a 28 d cycle). DNA from bone marrow (BM) and peripheral blood (PB) underwent bisulfite conversion and 1500ng of DNA was hybridized to Infinium Human Methylation27 BeadArrays. Arrays were processed by the Genome Sequencing Center at Washington University. The Infinium human Methylation27 Bead Array interrogates >27,000 well annotated CpG sites analyzing 14,500 promoters with single nucleotide resolution. Data was analyzed using BeadStudio version 3 and GraphPad Prism version 5. To compare the average percentages of methylated loci, we used a mixed model for repeated measurement data which allowed us to account for potential correlations among measurements taken from the same subjects. Loci were considered methylated if their average beta value was ≥0.80, hemi-methylated if their average beta value was <0.80 but >0.20, and unmethylated if it was ≤0.20. Loci on the X and Y chromosomes were excluded from analyses due to gender imbalances between responders and non-responders. Fifteen patients had BM and/or PB samples at baseline and/or on c2d25 and a response assessment. Median age was 69 years (range 53–82). IPSS categories were low (7%), Int-1 (27%), Int-2 (40%) and High (7%); FAB classifications were RA (26%), CMML (7%), RAEB (47%) and RAEB-t (20%). Best responses were: 2 patients with CR, 2 with a marrow CR, 1 with a PR, 8 with SD and 2 suffered PD. Patients with CR, mCR and PR were classified as responders whereas all others were classified as non-responders. For all patients the percentage of methylated loci decreased from baseline (13.4%) to c2d25 (9.9%) (p=0.0043). This effect was driven by a decrease in methylated loci in responders (15.1% to 8.9%) (p=0.0109). The decrease in methylated loci in non-responders was not statistically significant (12.4% to 10.5%) (p=0.1083). The difference in baseline methylation between responders and non-responders was not significant (p=0.1719). Baseline methylation was not associated with age (p=0.1475), IPSS (p=0.0934) or High- vs. Low-grade MDS (p=0.3575). Interestingly, the c2d25 assessment anteceded best clinical response in 4 out of 5 responders (all that achieved a mCR or CR) (81 days v. a mean of 202 days, SD 129 days to best response). Demethylation was likewise seen in the PB (18.6% to 11.7% of loci in responders and 18.8% to 13.6% of loci in non-responders) and overall there was good correlation between the mean beta values across the loci for BM and PB (r2=0.9937). But this correlation deteriorated when only methylated loci were included (r2=0.7936), demonstrating that the correlation is driven by hemi- or unmethylated loci, which compose ≥80% of loci, and suggesting that methylation in the PB is unlikely to be a good surrogate for BM methylation. Loci specific changes were considered potentially significant if average beta scores changed by ≥0.20 from baseline to c2d25. 11 loci met this criteria in responders and the changes in several of these were confirmed by t-tests (DNTT p=0.08; APOBEC2 p=0.0015, RGPD5 p=0.0138, TMSF4 p=0.0040). These differences and others between the baseline methylation status of loci in responders and non-responders will be correlated with mRNA expression levels (such as BCL2 p=0.08, HP p=0.0057, SHANK2 p=0.003). This study represents the first methylome-wide description of MDS and the changes induced by azacitidine. If validated in larger cohorts, BM methylation changes prior to cycle 3 may be able to help discriminate responders from non-responders earlier. Further studies are needed that include a larger sample size and that correlate changes in methylation status with changes in transcription and ultimately translation.

Topics:
azacitidine, dna, mineralocorticoid receptor, prostatic hypertrophy risk score, refractory anemia with excess blasts, bcl-2 protein, dna nucleotidylexotransferase, leukemia, myelomonocytic, chronic, myelodysplastic syndrome, nucleotides

Disclosures: Martin:Genzyme: Membership on an entity’s Board of Directors or advisory committees, Research Funding. Walgren:Eli Lilly and Company: Employment. Abboud:Genzyme: Consultancy; Novartis: Speakers Bureau; Pfizer: Membership on an entity’s Board of Directors or advisory committees; Bayer: Membership on an entity’s Board of Directors or advisory committees; Celgene: Membership on an entity’s Board of Directors or advisory committees. DiPersio:Genzyme: Honoraria; AnorMED: Honoraria; MGI Pharma: Honoraria. Vij:Pharmion: Research Funding, Speakers Bureau.

Author notes

Corresponding author

2008, The American Society of Hematology
2008
Sign in via your Institution