Role of epigenetic regulators in single-base substitution patterns in AML.

e19500 Background: Incidence of Acute Myeloid Leukemia (AML) increases with age, as does Clonal Hematopoiesis of indeterminate Potential (CHiP), wherein specific genes are mutated in healthy individuals. Single Base pair Substitutions (SBS) are associated with aging and AML, most commonly C>T. G>A transitions are found in a similar prevalence in all age groups, although the number of mutations are higher in older age groups and associated with therapy. Determining a pattern of SBS in leukemia that is distinct from aging has been a challenge. Methods: We performed mutational signature analysis on 10,152 mutations in 3,575 unique suspected AML patient bone marrow, peripheral blood, or FFPE tissue samples. Samples were evaluated using an all exon amplicon based 27 gene next generation sequencing panel. The cohort was divided into 5 groups based on age (0-14, 15-39, 40-54, 55-65, and >65 years), as well as into three gene mutation groups 1) TET2 WT (n=7,479 mutations), 2) TET2/DNMT3A WT (n=6,701) and 3) TET2/DNMT3A WT with chromatin regulators WT (ASXL1, ATRX, BCOR, BCORL1, EZH2, PHF6, RAD21, SMC1A, SMC3, STAG2; n=3627 mutations). Results: Predominant SBS were C>T (42.7%) and G>A (59.0%) transitions across all age groups. The percentage of C>T SBS increased with age: 0-14 years cohort (18.0% of cases; 1.5% of mutations) to >65 years cohort (58.8% of cases; 39.6% of mutations). There was no significant change in the prevalence of G>A mutations with age, however, the total number of mutations did increase with age (0-14 0.2% vs >65 39.6%). In each of the gene mutation groups, the number of C>T and G>A total mutations decreases as TET2 MT (42.7% and 59.0%), TET2 WT (27.0% and 45.1%), TET2/DNMT3A WT (20.8% and 34.3%), and TET2/DNMT3A WT with WT chromatin regulators (10.1% and 25.5%). Conclusions: In this study, both the prevalence and number of mutations in AML patients increases with age for C>T transitions, but for G>A transitions only the number of mutations increases with age, consistent with previously reported studies. In this data set, by removing mutations in TET2, TET2 and DNMT3A, TET2 and DNMT3A with genes that regulate chromatin accessibility, the number of C>T and G>A transitions decreases in order of effect, with the chromatin modifiers having the largest effect. The effect of removing these mutations suggests a mechanism driving leukemogenesis whereby DNMT3A methylates cytosine, TET2 deaminates this methylated cytosine, and finally chromatin modifiers regulate accessibility to these methylation sites, thereby modulating downstream oncologic pathways.