CpG Island Methylator Phenotype Of Myelodysplastic Syndrome Identified Through Genome-wide Profiling Of DNA Methylation And Gene Expression And Its Clinical Significance

Background and objective:The CpG island methylator phenotype (CIMP) is an epigenetic phenomenon and plays an important role in the tumorigenesis in a variety of cancers. The identification of aberrant DNA methylation can be exploited for early diagnosis, treatment planning, and risk assessment of cancer patients. The objective of this study was to identify CIMP in myelodysplastic syndrome (MDS) and a combination of methylated marker genes to improve the diagnosis and prognosis of MDS.Methods:Genes were screened for hypermethylation through genome-wide DNA methylation profiling. Transcription downregulation was determined with a gene expression microarray. Methylation-specific, real-time, and bisulfite-sequencing PCR cloning and sequencing were performed to validate selected genes. The hypermethylation of genes as a diagnostic tool for the detection of MDS was evaluated singly and in combination. The Kaplan-Meier test was performed to calculate and generate survival curves. Comparisons of OS or LFS between groups were evaluated by the log-rank test. The Cox regression model was used for multivariate survival analysis to identify the significant independent prognostic factors affecting OS or LFS.Results:A draft of an MDS CIMP was established and revised to 6 genes after validation in 20 patients and 20 controls. Further large-scale analysis showed that, of 211 MDS patients, the majority were hypermethylated in ABAT (97%), DAPP1 (98%), FADD (89%), LRRFIP1 (96%), PLBD1 (89%), and SMPD3 (85%). The area under the curve (95% CI) of CIMP was 0.9768 (0.9609-0.9928). A combination of 5 or more of the methylated genes showed a specificity of 95% and sensitivity of 91% for the diagnosis of MDS. We thus defined patients with 5 or more of the methylated genesas CIMP-positive. To further assess the influence of CIMP status on LFS and OS, we performed a Kaplan-Meier survival analysis and found that being CIMP-positive was a significant negative prognostic factor. The differences in median survival time(P= 0.0308) and overall survival time(P= 0.0422) between the CIMP-positive group and CIMP-negative group were significant. We next analyzed OS and LFS by univariate analysis for the clinical and hematological factors and CIMP. And the following factors had a significant impact on both OS and LFS:CIMP status, age, hemoglobin level, mean corpuscular volume (MCV), WHO classification, marrow blast level, karyotypes, cytopenias defined by the IPSS, and the IPSS risk group itself. To assess the prognostic value of factors found to be significant in the univariate analysis, we performed a multivariate Cox regression analysis of the effect of the hemoglobin level, age, MCV, marrow blast levels, karyotypes, and cytopenias defined by IPSS, and the CIMP status on OS. Results of the multivariate analysis indicated that CIMP status had the strongest influence on OS (hazard ratio= 4.82), indicating that the CIMP was an independent factor affecting OS.Conclusion:We demonstrated here that the genes ABAT, DAPP1, FADD, LRRFIP1, PLBD1 and SMPD3 were hypermethylated and downregulated in MDS. And we defined these six genes as CIMP markers of MDS, which could be a noninvasive approach for the diagnosis of MDS, and indicating prognosis.