Clonal Evolution and Relapse in Adult De Novo Acute Myeloid Leukemia

Patients with intermediate risk AML have unpredictable outcomes. Some do extremely well with conventional induction and consolidation chemotherapy, while others do poorly. The discovery of nearly all mutations associated with AML pathogenesis created an important database that has allowed investigators from around the world to use panel tests to assess the mutational landscapes of thousands of patients with known outcomes. Although these studies have been extremely informative, they have not dramatically altered the ability to accurately predict whether an individual intermediate-risk AML patient will relapse. The reasons are multifold. First, there is enormous combinatorial complexity of the mutations associated with AML. At least three mutations are required for the development of most AML cases, but since there are more than 250 recurrently mutated genes, the combinatorial possibilities are enormous. Secondly, because virtually all patients with AML present with clonal heterogeneity, and since individual subclones have different susceptibilities to chemotherapy, it has been difficult to predict outcomes based solely on the mutational composition of a patient at presentation. Further, since most subclones are 'invisible' to flow cytometric methods, and do not contain well-established mutations in genes that are recurrently mutated in AML, we cannot yet predict whether a subclone may be refractory to chemotherapy, or have extraordinary relapse potential . To overcome these issues, we asked whether information based on the clearance of the subclones from each patient could "integrate" this critical information in a practical way, and ultimately improve risk assessment. This study (Klco et al, JAMA 2015) suggested that a practical evaluation of mutation clearance after induction therapy could represent an improvement in our ability to predict whether intermediate risk patients who fail to clear their mutations after induction would benefit from allogeneic transplantation as their primary consolidation. A prospective clinical trial designed to answer this question is underway. Efforts to identify the properties of subclones that have high relapse potential, using single cell RNA-seq, and the sequencing of the relapse genomes of additional AML cases, will also be discussed.