Myelodysplastic Syndrome Progression to Acute Myeloid Leukemia at the Stem Cell Level

Abstract

Myelodysplastic syndromes (MDS) are malignant hematopoietic stem cell-derived disease with poor clinical outcome and overall survival. Overall ~30% of patients with MDS progress to secondary acute myeloid leukemia (sAML), which has an exacerbated median survival of less than six months. Delaying the progression to sAML represents one of the key challenges in the clinical management of patients with MDS. Thus, cellular and molecular insights into the progression of MDS to sAML are needed. Previous seminal studies have investigated the clonal evolution during MDS to sAML progression by sequencing the longitudinal samples of bulk tumor cells from MDS and matched sAML of the same patient. However, the clonal origin of MDS and AML has been demonstrated to lie within the phenotypically and functionally defined stem cell compartment, representing a small subset of total bone marrow which cannot be effectively interrogated by bulk sequencing. Clonal evolution at the stem cell level has not yet been directly examined.

To prospectively analyze subclonal composition at the stem cell level in MDS and study their clonal evolution during progression to sAML, we utilized longitudinal, paired samples from 7 patients with MDS who had later progressed to sAML. For both MDS and paired sAML samples, we utilized multi-parameter fluorescence-activated cell sorting (FACS) to isolate phenotypically defined stem cells and blasts from both stages. Specifically, we isolated hematopoietic stem and progenitor cells (HSPC, Lineage -CD34+CD38 -) expressing at least one of the leukemic stem cell (LSC) markers (CD45RA, CD123, or ILIRAP), to enrich for malignant stem cells (MDS-SC, AML-SC). At the same time, we isolated HSPCs that were triple-negative for CD45RA, CD123, and ILIRAP to enrich for pre-malignant stem cells (preMDS-SC, preAML-SC). We observed that phenotypically defined malignant stem cells were significantly expanded in the sAML compared to the matched MDS stage. Moreover, compared to preMDS-SC and preAML-SC, sorted MDS-SCs and AML-SCs showed impaired differentiation, with significantly myeloid-biased outputs in vivo and ex vivo, demonstrating that CD45RA/CD123/ILIRAP expressing HSPCs are indeed enriched for malignant stem cells in MDS and AML.

To investigate the patterns of somatic mutations in stem cells and blasts, we performed targeted deep sequencing covering both coding and non-coding regulatory regions of around 600 MDS and AML relevant genes. Interestingly, we found that stem cells at both MDS and sAML stages have a significantly higher number of subclonal mutations compared to the blast population. Mutation signature analysis revealed that mutation patterns in all the cell populations were mainly associated with age-related signatures. Interestingly, in stem cell compartments, there are specific mutations associated with DNA repair pathways. We next directly compared the subclonal diversity of stem cells versus blasts by clonality analysis, which also revealed higher subclonal complexity at the stem cell populations compared to the blasts. In addition, in all the 7 patients studied, we observed shared dominant clones across all stem cell and blast populations, with key mutations such as TET2, TP53, or U2AF1. Remarkably, these shared mutations were also detectable in the T cells isolated from the same patients, suggesting that these mutations in TET2, TP53, and U2AF1 were acquired at the early stage during the initiation of MDS and AML.

We next performed single-cell targeted re-sequencing of sorted stem and blast cells, to experimentally examine the clonal architecture and evolution during MDS to sAML progression. Our results revealed a pattern of non-linear, parallel clonal evolution, with distinct subclones within the MDS stem cells contributing to the generation of MDS blasts or the progression to sAML, respectively. The progression to sAML can be associated with rare subclones in MDS stem cells harboring mutations on RUNX1, NRAS, or NTRK3 and DUSP22 in 3 out of 7 patients, or a larger subclone with ERG and ATRX mutations in one patient. These MDS stem cell subclones were not detectable in MDS blasts, but became dominant upon sAML progression, indicating the early branching evolution at the MDS stem cells during the blast generation and disease progression to sAML. Additionally, in 3 other patients, we observed slightly later branching evolution at MDS stem cells, with dominant clones shared across MDS and sAML stages.

In summary, using sorted stern cell populations from patients with MDS, we were able to capture of even small subclones at MDS stem cells that are crucial for eventual progression to sAML. Our data provide a comprehensive characterization of subclonal diversity, with important insights into the subclonal evolution of stem cells in MDS pathogenesis and transformation, as well as have implications for current bulk cell-focused precision oncology approaches.