Clonal Analysis of Acute Myeloid Leukemia Secondary to Myeloproliferative Neoplasms

[Background] Prognosis of patients with secondary acute myeloid leukemia (sAML) transformed from myeloproliferative neoplasms (MPNs) is extreamly poor. It has been demonstrated that several additional genetic alterations, which are engaged in epigenetic regulation, cell-growth signaling and RNA splicing machinery, cooperate with JAK2, CALR and MPL mutations for the development of sAML from MPN. The order of mutation acquisition also influences the clonal evolution in MPN. However, little is known about the mechanisms of its leukemic transformation. In previous studies, both JAK2 -mutated and -unmutated AML clones were identified in sAML cases transformed from JAK2 -mutated MPNs, suggesting the diversity of the clonal evolution during the leukemic transformation from MPNs. In this study, we investigated genetic alterations and clonal evolutions in sAML from MPN, especially in JAK2 -unmutated sAML, to elucidate the mechanisms of the leukemic transformation.

[Methods] We performed targeted deep sequencing of 54 genes in six paired MPN and sAML samples. Whole-exome sequencing (WES) was also performed in a paired MPN, sAML and normal tissue cells. For a clonal analysis, bone marrow (BM) cells at complete remission (CR) stage after chemotherapy were sorted into hematopoietic stem cell (HSC) and hematopoietic progenitor fractions such as common myeloid progenitor (CMP), granulocyte-macrophage progenitor (GMP) and megakaryocytic/erythroid progenitor (MEP) fractions, and analyzed the mutations detected in targeted deep sequencing or WES. Furthermore, we sorted single cells from the HSC fraction and performed single-cell mutational analysis to clarify the clonal evolutions precisely during the development of sAML. In addition, we intravenously inoculated BM cells at the sAML stage into NOD/SCID/IL-2Rγnull (NOG) mice, and compared mutation status between primary AML cells and engrafted leukemia cells to evaluate the clonal hierarchy of leukemia cells.

[Results] Four JAK2 -mutated and two CALR -mutated MPNs were analyzed in this study. Targeted deep sequencing identified TP53 mutations in five of six sAML samples with high variant allele frequencies (VAFs). In two of the four JAK2 -mutated MPNs, VAFs of JAK2 -V617F were significantly decreased in sAML cells compared with those in MPN cells, whereas VAFs of TP53 mutations were emerged or increased in sAML cells, suggesting that sAML clones with TP53 mutations were raised from JAK2-V617F negative hematopoietic cells in both cases. Moreover, we identified the mutations which was common to MPN and AML cells in both cases. These mutations formed common initiating clones prior to MPN and sAML, and additional mutation in JAK2 or TP53 was associated with their MPN or sAML development. The single-cell analysis in CD34⁺/CD38^- HSC fractions revealed that a small part of the cells harbored TP53 -R248W and JAK2 -wild type (WT), half of the cells harbored homozygous JAK2 -V617F and TP53 -WT, and the other half cells harbored JAK2 - and TP53 -WT. Notably, three common mutations detected both in MPN and in sAML with the same VAFs were mainly found in JAK2 -WT cells. Thus, we validated that JAK2 and TP53 mutations were mutually exclusive HSC fractions, and confirmed that sAML clones were raised from JAK2 -unmutated clone. In patient derived xenotransplant models using JAK2 -WT sAML cells following JAK2 -mutated MPN or CALR -mutated sAML, engrafted leukemia cells showed the same mutational pattern as primary sAML cells. These results suggested that sAML clones but not MPN clones were able to engraft and proliferate in mice.

[Conclusions] The genetic analysis in paired MPN and sAML samples revealed the diverse clonal evolution patterns during the development of sAMLs from MPNs. TP53 mutation was strongly associated with leukemic transformations both in JAK2 -V617F mutated and CALR -mutated MPNs. However, we demonstrated two cases in whom driver mutations in MPNs were not preserved in sAML clones. In those cases, there might be independent common initiating clones corresponding to MPN or sAML, and each initiating clone might be acquired JAK2 or TP53 mutation before the development of MPN or sAML, respectively. Our data also suggested the importance of single-cell analysis which enable us to clarify the clonal evolutions in leukemic transformation from MPN, that were not identified in targeted deep sequencing as well as WES.