Biology of Myelodysplasia Clonality: Clinical Implications of Whole-Genome Sequencing

Abstract SCI-32

The discovery of acquired genetic mutations associated with myelodysplastic syndromes (MDS) and acute myeloid leukemia (AML) has rapidly expanded with the use of next-generation sequencing technologies. The distinction between MDS and de novo or secondary AML (arising after MDS) and therapeutic recommendations are largely based on the presence of morphologic dysplasia and the percentage of bone marrow myeloblasts. Using whole genome sequencing to define the allele burden of hundreds of mutations, our group has shown that nearly all the bone marrow cells in seven patients with MDS and secondary AML were clonally derived (∼85%), regardless of the myeloblast count. We now have experimental evidence suggesting that the vast majority of mutations in these samples are randomly acquired during aging (passenger mutations) and carried forward in a cell when it is transformed. By using all the validated mutations in a sample (passengers and drivers) we found that five of the seven MDS samples harbored two or more malignant clones and that all seven contained a founding clone that gave rise to subsequent daughter clones. Thus, the MDS bone marrow samples are a mosaic of several tumors with unique sets of mutations each shaped by the acquisition of a series of mutations and clonal diversification. Each clone contained at least one coding gene mutation, and we identified 11 recurrently mutated genes in the seven samples, including recently described mutations in U2AF1, a spliceosome gene that is typically mutated in the MDS founding clone. In addition, the MDS founding clone always gave rise to the dominant secondary AML clone, suggesting that therapies targeting mutations in the founding clone may have benefit in treating both MDS and secondary AML. It is possible that therapeutic response and disease progression is driven not only by the presence of recurrent mutations, which have prognostic value, but also by the clone (i.e., founding and daughter) in which they arise. We are currently testing this hypothesis by monitoring the dynamic change in clone size, including driver mutation allele burdens, during treatment of patients with AML and MDS.