Research Highlights: Highlights from the Latest Articles in Epigenomics

Abstract

Shinichi Fukushige*,1,2 & Akira Horii1 Department of Molecular Pathology, Tohoku University Graduate School of Medicine, Sendai, Miyagi 980–8575, Japan Center for Regulatory Epigenome & Diseases, Tohoku University Graduate School of Medicine, Sendai, Miyagi 980–8575, Japan *Author for correspondence: Tel.: +81 22 717 8043 Fax: +81 22 717 8047 fukushige@med.tohoku.ac.jp A seven-gene score comprising genetic & epigenetic information has a prognostic impact in AML Evaluation of: Marcucci G, Yan P, Maharry K et al. Epigenetics meets genetics in acute myeloid leukemia: clinical impact of a novel seven-gene score. J. Clin. Oncol. 32(6), 548–556 (2014). Nearly half of acute myeloid leukemia (AML) patients harbor normal karyotype, termed cytogenetically normal (CN)-AML. Prognosis of CN-AML vary, presumably due to different genetic and/or epigenetic alterations. Therefore, it is important to establish specific and accurate biomarkers for better classification of CN-AML. Marcucci et al. have used epigenetic and genetic prognostic information in AML, and have developed a clinically meaningful scoring system [1]. The authors first used a training set of older (≥60 years) patients with CN-AML (n = 134) and identified 32,681 unique differentially methylated regions (DMRs) by comparing mutated versus wild-type patients for each of the most common prognostic 15 mutations in AML. The authors further identified 82 individual genes whose promoter DMRs were associated with overall survival (OS); among these, higher DNA methylation as well as lower expression levels were significantly associated with longer OS in seven genes (CD34, RHOC, SCRN1, F2RL1, FAM92A1, MIR155HG and VWA8). A weighted summary score of the expression levels of these seven genes was analyzed for its prognostic significance in the training set as well as the four validation sets. In total, low-score patients, compared with high-score patients, had a higher complete remission (CR) rate (85 vs 58%, respectively; p < 0.001) and longer disease-free survival (DFS) (3 year DFS, 40 vs 15%, respectively; p < 0.001) and OS (3 year OS, 48 vs 15%, respectively; p < 0.001). Younger (<60 years) and older patients with one gene or fewer with high expression had the best outcomes (CR rate, 94 and 87%, respectively; 3 year OS, 80 and 42%, respectively). To date, risk assessment for patients with AML has primarily relied on cytogenetic and genetic testing. Shen et al. have analyzed the prognostic significances of mutations of the 12 most frequently mutated genes in AML in 605 AML patients with CN or 11q23 abnormality [2]. In these patients, DNMT3A and MLL mutations were poor prognostic factors, whereas biallelic CEBPA mutations or NPM1 mutations without DNMT3A mutations conferred a better outcome. However, it was not easy to make a standardized prognostic algorithm based on mutation patterns. Even though aberrant DNA methylation in AML has been studied extensively throughout the genome, clinically applicable prognostic biomarkers are still not available [3]. A new approach to integrate genetic and epigenetic information is promising and in fact identified clinically meaningful groups of patients with AML who responded differently to chemotherapy. Recently, a number of epigenetic regulator genes, including TET2, IDH1, IDH2, DNMT3A, ASXL1 and EZH2 frequently mutated in AML, has also been mutated in many types of cancers [4]. For example, IDH1 mutation appears to be a causal contributor to the CpG island methylator phenotype (CIMP) in glioblastoma and CIMP is associated with a favorable prognosis [5]. The idea presented in this article might be applicable and useful for risk stratification and prognostication in many types of cancer. Research Highlights