Background: Acute myeloid leukemia (AML) accounts for the most leukemia-related deaths in the United States and its incidence has been rising as the population ages. Although certain molecular aberrations are prognostic and have come into mainstream clinical practice, the genetic and epigenetic determinants of curability in AML remain incompletely understood. Our study examines the role of DNA methylation patterns in AML prognosis and expands on our preliminary work showing DNA hypermethylation may associate with improved overall survival.

Methods: To quantitatively interrogate genome-wide CpG methylation we used Digital Restriction Enzyme Analysis of Methylation (DREAM) on a cohort of 102 AML patient samples and 25 normal control samples. We validated our findings using DNA methylation data from 194 patient samples from The Cancer Genome Atlas (TCGA) on the Illumina Infinium HumanMethylation450 platform. Statistical analysis was done using R.

Results: Preliminary analysis by our group of DNA methylation levels at promoter CpG islands (CGI) of OSCP1, NPM2, OLIG2, SCGB3A1, and SLC26A4 showed significant hypermethylation in a small group of long-surviving AML patients compared to a short-surviving cohort (median OS = 2,694 days vs. 207 days). We expanded on this observation using DREAM to measure genome-wide DNA methylation in clinical AML samples and found that hierarchical clustering based on 2,537 CpG sites with a standard deviation above 20% stratified patients into three groups with significant differences in overall survival. The hypermethylated cluster had the best prognosis and seemed to be defined by hypermethylation at promoter CGIs, suggesting that a CGI methylator phenotype (CIMP) in AML may be a favorable prognostic factor (median OS: CIMP = 5,110 days, Cluster 2 = 380 days, Cluster 3 = 555 days; log-rank p=0.0162). We then validated these findings using TCGA data from AML patient samples. Hierarchical clustering on the basis of CGI promoter sites revealed three distinct groups with the CIMP cluster having significantly improved overall survival compared to the other clusters (median OS: CIMP = 761 days, Cluster 2 = 306 days, Cluster 3 = 365 days; log-rank p=0.0013). There was also a trend in overall survival when non-CGI non-promoter sites were used to cluster samples (median OS: CIMP = 593 days, Cluster 2 = 245 days, Cluster 3 = 456 days; log-rank p=0.1530). Consistent with our DREAM data, combining CGI promoter sites with non-CGI, non-promoter sites revealed a hierarchical clustering pattern of three major clusters with significant differences in overall survival (median OS: CIMP = 822 days, Cluster 2 = 365 days, Cluster 3 = 365 days; log-rank p=0.0295). Despite technical differences between platforms, there was significant overlap in the genes most proximal to differentially methylated sites between the DREAM and TCGA analyses. These common genes were significantly enriched in transcription factors, pyrimidine metabolism genes, and development genes. Interestingly, the presence of IDH1 R140 mutations was significantly greater in the CIMP clusters in both the DREAM and TCGA analyses (p<0.0001, p=0.0004 for DREAM and TCGA respectively). Taken together the results from our DREAM data and the TCGA validation support the hypothesis that CIMP is a favorable prognostic feature in AML.

Conclusions: We propose that the CIMP methylation pattern is associated with favorable prognosis in AML. We have identified a subset of methylation sites that, when interrogated, predict overall survival independent of other clinical factors.

Citation Format: Andrew D. Kelly, Heike Kroeger, Jumpei Yamazaki, Rodolphe Taby, Frank Neumann, Justin T. Lee, Rong He, Shoudan Liang, Yue Lu, Matteo Cesaroni, Sherry A. Pierce, Steven M. Kornblau, Carlos E. Bueso-Ramos, Farhad Ravandi, Hagop M. Kantarjian, Jean-Pierre J. Issa, Jaroslav Jelinek. Genome-wide methylation analysis reveals an independently validated CpG island methylator phenotype associated with favorable prognosis in acute myeloid leukemia. [abstract]. In: Proceedings of the AACR Special Conference on Hematologic Malignancies: Translating Discoveries to Novel Therapies; Sep 20-23, 2014; Philadelphia, PA. Philadelphia (PA): AACR; Clin Cancer Res 2015;21(17 Suppl):Abstract nr B22.