MDS Progression to AML at the Stem Cell Level

Myelodysplastic syndromes (MDS) frequently progresses to acute myeloid leukemia (AML), however, the cells leading to malignant transformation have not been directly elucidated. As progression of MDS to AML in humans provides a system to determine the cellular origins and mechanisms of neoplastic transformation, we studied stem cell populations as well as blasts in longitudinal samples of patients with MDS who progressed to AML. Highly fractionated (stem) cell populations were subjected to targeted deep sequencing (customized panel of ~500 genes) resulting in unprecedented sequencing depth at the stem cell level (equivalent to ~500,000x to 2,500,000x at the bulk level). Unexpectedly, we found that stem cell populations at both the MDS and AML stages had a significantly higher number of mutations compared to blasts (p < 0.01). Clonal analysis also showed higher subclonal complexity at the stem cell level compared to blasts (p < 0.05). Strikingly, deep sequencing combined with single-cell targeted re-sequencing of sorted stem cells revealed a process of non-linear, and rather parallel clonal evolution at the stem cell level during initiation of MDS and progression to AML in all patients: the mutational process generated numerous distinct subclones at the stem cell level, and only one or few of these clones become apparent at the bulk/blast level. Furthermore, we observed different subclones within the MDS stem cells contributed to the generation of MDS blast, or the progression to AML, respectively. Rare subclones already present in MDS stem cells (e.g. harboring mutations in RUNX1, NTRK3, and DUSP22), or larger subclones (e.g. with ERG and ATRX mutations), were not detectable in MDS blasts but became dominant upon AML progression. Taken together, our results suggest a fundamentally novel model of malignant transformation at the stem cell level, and reveal a crucial role of highly diverse stem cell compartments during MDS progression to AML. In addition, these findings have significant implications for current bulk cell-focused precision oncology approaches in MDS that may miss important leukemia-driving stem cell subsets and mutations.