Towards an inclusive genomic risk classification for acute myeloid leukemia (AML)

Acute myeloid leukemia (AML) is the most common adult acute leukemia. Molecular features such as cytogenetics and somatic mutations are essential components of risk stratification; used in daily clinical practice to determine treatment modality and intensity. Our knowledge of recurrent genetic AML-associated features, survival associations, subsequent genetic risk classification and clinical practice is informed by large-scale genomic studies performed over the past decade. Current clinical practice is predicated on the supposition that adequate testing has been performed and that genetic background should not interplay with the known AML-associated genetic and genomic landscape or with genes that associate with treatment response and/or that drive AML-genesis. However, our published and preliminary data reveal that constitutional genetic elements affect not only the frequencies and impact of known AML-associated gene mutations, but also multiple recurrent variants in genes thus-far not implicated in AML-genesis. We hypothesize that: a) known AML-associated drivers may carry different prognostic significance and might need to be adjusted in clinical risk classifications depending on genetic backgrounds, and b) some unrecognized molecular features are novel drivers of AML-genesis and/or therapy resistance. We propose to extend our initial studies of AML genomes and genomics to provide statistical confidence around frequently mutated genes, and a more precise, molecularly-adjusted risk and treatment stratification for AML patients. Next, we will focus on NPM1c, a genotype which currently is predicted to confer favorable outcome in AML patients. We will use cutting-edge single-cell multiomic assays to delineate bystander clonal mutations from potential driver mutations, then we will biologically test their roles in clonality/leukemia-stem-cell frequency and treatment response in genetically heterogeneous patients. We predict this approach to provide an improved characterization of the genetic and genomic landscape of AML, identify those variants with prognostic significance, and provide exemplars of heretofore unrecognized drivers of treatment response and survival. The overall goal is to enable precision oncology approaches which accommodate the effects of underlying human genetics in all patients.