Genomic Landscape of Acute Myeloid Leukemia (AML) on the Basis of 2017 ELN Classification and Other Mutations in Adult AML - Single Healthcare System Data

Abstract

Introduction Recently, genomic mutation profiling of leukemic cells has been actively studied and some results have been integrated into the 2017 ELN classification with cytogenetic analysis for risk assessment of AML populations.1 However, only a few mutations are well identified and included in the 2017 ELN classification. In addition, except interaction between NPM1 and FLT-ITD, correlation and co-occurrence among various mutations have not been well studied. Here we describe our single center genomic landscapes of 2017 ELN guideline components with other NGS mutations in adult AML. Methodology We performed hematopoietic tumor profiling assay by next generation sequencing (NGS) testing and cytogenetic chromosome analyses in 193 AML patients diagnosed from 2018-08-01 to 2020-03-10. Treatment-related AML patients were excluded. All NGS and cytogenetic analyses were performed before starting chemotherapy. On the basis on 2017 ELN cytogenetic and mutations components, we analyzed significantly other co-occurred and exclusively occurred NGS mutations by using Fischer's exact test. Based on 2017 ELN classification, either one of t(6;9), t(v;11q23.3), t(9;22), inv(3) or t(3;3),-5, del(5q), -7, -17, or complex/monsoonal karyotype meeting the ELN criteria positive was grouped as Adverse Karyotype. 43 AML related genes, including 6 ELN components of NPM1, FLT-ITD low, CEBPA biallelic, TP53, RUNX1, ASXL1 and other 37 mutations including DNMT3A, NRAS and KRAS were analyzed as a single component. In addition, we grouped some mutations into larger sets or pathways and analyzed them in the same way. DNMT3A, TET2, IDH1, IDH2, and SETBP1 were grouped as DNA methylation. SFSB1, SRSF2, U2AF1, and ZRSR2 were grouped as spliceosome. BCOR, CBORL1, EXH2, and KDM6A were grouped as chromatin modifier. ASXL1, which is also chromatin modifier, was analyzed as a separated component because it is a part of the 2017 ELN. NRAS and KRAS were analyzed separately and also as one group. One-tailed statistical significance is at level of 5% for statistical analysis. Results Our cohort was male predominant (57%, 110/193) with median age of 64 YO (range 18 - 93). 29%, 25% and 46% of patients were 2017 ELN favorable, intermediate, and adverse group respectively. 27.5% and 40% were Adverse and Normal Karyotype respectively. 4%, 4% and 1% of patients were RUNX1-RUNX1T1, CBFB-MYH11 and MLLT3-KMT2A positive respectively. Figure1 describes the occurrence rate of significantly occurred mutations. Regarding 2017 ELN components, 37 patients (19.2%) were positive to NPM1, 34 patients (17.6%) had FLT-ITD low, 3 (1.6%) had CEBPA biallelic, 40 (20.7%) had TP53, 27 (14.0%) had ASXL, and 24 (12.4%) had RUNX. There was no FLT-ITD high mutation in our cohort. In total 63 patients (32.6%) had at least one DNA methylation mutation. 44 patients (22.8%) had at least one spliceosome mutation and 9 patients (4.6%) had at least one chromatin modifier mutation other than ASXL1. In activated signaling, NRAS, KRAS, FLT-TKD were significantly occurred. Table 1 describes significantly co-occurred and exclusively occurred mutations and Figure 2 visualizes significantly co-occurred and exclusively occurred mutations. TP53 exclusively occurred with normal karyotype and many different mutations including NPM1, FLT-ITD low, DNA methylation group, and KRAS/NRAS, and significantly co-occurred with Adverse Karyotype. NPM1 significantly co-occurred with Normal Karyotype and FLT-ITD mutations. Interestingly, DNA methylation group and especially IDH2 and DNMT3A significantly co-occurred with NPM1. Also, RUNX1, ASXL1 and Spliceosome group co-occurred with each other. KRAS/NRAS co-occurred with CBFB-MYH1. KIT co-occurred with CBFB-MYH1 and RUNX-RUNX1T1, although the sample sizes were relatively small. Discussion Based on strong accumulated evidence, ELN established new guidelines in 2017. Therefore, it would be a sophisticated way to find other significant mutations and their interactions on the basis of already well established 2017 ELN components. Our findings suggest DNA methylation regulatory genes, especially DNMT3A and IDH2, significantly co-occurred with NPM1, which is a component of the favorable group. Also, spliceosome mutations significantly co-occurred with RUNX1 and ASXL1, which are components of the adverse group. Further research is needed to determine whether those cooccurrences affect each group's prognosis. Figure Disclosures No relevant conflicts of interest to declare.