Cytopenic complications are the major cause of morbidity and mortality in patients with myelodysplastic syndromes (MDS), and alleviation of cytopenias is an important goal of treatment in any patient diagnosed with MDS. However, MDS is not a static disease: Approximately one-third of patients progress to secondary acute myeloid leukemia (sAML), signified by a maturation arrest of the malignant clone resulting in an increase in blasts in the blood and/or bone marrow (at least 20%) or an extramedullary blast proliferation (myeloid sarcoma). Development of sAML is an adverse event that is associated with resistant disease and typically, short survival. Moreover, sAML progression is often used as an end point in clinical trials that evaluate treatment responses and as a surrogate for overall survival in prognostic models. The same features that predict shorter survival in MDS — worse cytopenias, higher bone marrow blast percentages, and adverse genetic risk — are also associated with higher likelihood of progression to sAML.1, 2 However, while it is known that mutations sequentially accumulate along the trajectory between MDS and sAML,3, 4 the details of this process are still poorly understood. In particular, the role of expansion of pre-existing clones versus the emergence of new clones and the influence of the clonal architecture (pattern of subclones with distinct mutation portfolios) on disease progression need further elucidation.
Dr. Andrew J. Menssen and colleagues have studied the dynamic clonal architecture in a series of 43 patients with MDS who progressed to sAML. They used not only bulk next-generation sequencing (NGS) panels, but also whole-genome sequencing, single-cell sequencing, and highly sensitive error-corrected NGS on subsets of cases to resolve clonal architecture at the cellular level and achieve deeper sensitivity to detect low-level mutations. The authors found that mutations in transcription factor genes (such as ETV6, GATA12, and RUNX1) and in signaling genes (including mostly RAS pathway genes such as CBL, FLT3, KRAS/NRAS, and PTPN11) are typically subclonal in MDS, occurring over a “backbone” founding clone characterized by epigenetic modifier gene mutations (e.g., TET2 or EZH2) and often accompanied by spliceosome gene mutations (e.g., SF3B1), as was concurrently reported by Dr. Tiffany Guess and colleagues.5 Although signaling gene mutations detected by standard NGS panels were relatively infrequent in the MDS phase of disease, by sensitive PCR methods, low-level subclonal signaling gene mutations (at variant-allele fraction <2%) were detected in a subset of patients. They found diverse and variable patterns of evolution in the clonal architecture as MDS progressed to AML, with mutations waxing and waning throughout the disease course. In many cases, new signaling gene mutations arose concurrent with progression to sAML, while in other cases pre-existing subclonal signaling mutations expanded or contracted as the blast count increased (Figure). The authors extended their observations to a separate cohort of 135 patients with MDS to perform sensitive error-corrected NGS and associate the findings with progression to sAML. They found that, in addition to the expected association of higher Revised International Prognostic Scoring System (IPSS-R) risk with progression to sAML (p=0.023), the presence of a signaling gene mutation was also independently associated with progression (p=0.007). Interestingly, this predictive effect of signaling pathway mutations for sAML progression was limited to lower-risk (IPSS-R very low/low/intermediate) patients, and not the IPSS-R higher-risk patients.
![Single-cell sequencing of a secondary acute myeloid leukemia (sAML) case from the study of Dr. Menssen and colleagues illustrates how the dynamic clonal architecture can be inferred by the subclonal mutation patterns. A) An initial clone carrying SRSF2 (spliceosome) and ASXL1 (epigenetic) mutations subsequently acquired a RUNX1 mutation (transcription) and then a PTPN11 mutation (signaling) as the final dominant clone in the sAML; a smaller subclone instead acquired NRAS, another signaling mutation. B) Complex clonal evolution pathways of signaling gene mutations in the progression from MDS to sAML. While a large subset of patients (26%) acquired a new signaling gene mutation upon development of sAML, significant subsets had pre-existing mutations that expanded (7%) or contracted (9%). Complex patterns were also observed (lower four panels) in which some subclones persisted simultaneously with the acquisition or loss of signaling mutations in other clones. (Adapted from Figure 3A and Figure 5D from Menssen AJ. Convergent clonal evolution of signaling gene mutations is a hallmark of myelodysplastic syndrome progression. Blood Cancer Discov. 2022;3[4]:330-345.)](https://ash.silverchair-cdn.com/ash/content_public/journal/thehematologist/19/6/10.1182_hem.v19.6.202263/3/m_hem190603f1.jpeg?Expires=1720784948&Signature=mLcnbAyBW3eDpG1P95LAfaMY4mQDdUGN9B-m60cBCGAKtuH6EgzPSo4ewI7J~fRi3Hlpkk-88y2Gd84zaSFNVpm~my1K5Cvuo5NivABlMkC37bsLBG1Iwb~aFlZG-v0SOc4re6GXvIAUbCCn3sbwyP~IONvaf~PBmj6IehR7~y3kSTB0H3BZB24~5JeJFMosR1XaSqFTVUMC~UF3g2y6hvFnANGSIFRQWjNqqgl1MNoPDbX33AkiHBK7wBqilqCKZ42E7F2z9hnDwXVFmqlnfVKoNSbCBi4~v4ZkpebIVW0SjFafsYasQ1PV78KeH5ENXLO1RCS0tnh1MX1lqyxROQ__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
Single-cell sequencing of a secondary acute myeloid leukemia (sAML) case from the study of Dr. Menssen and colleagues illustrates how the dynamic clonal architecture can be inferred by the subclonal mutation patterns. A) An initial clone carrying SRSF2 (spliceosome) and ASXL1 (epigenetic) mutations subsequently acquired a RUNX1 mutation (transcription) and then a PTPN11 mutation (signaling) as the final dominant clone in the sAML; a smaller subclone instead acquired NRAS, another signaling mutation. B) Complex clonal evolution pathways of signaling gene mutations in the progression from MDS to sAML. While a large subset of patients (26%) acquired a new signaling gene mutation upon development of sAML, significant subsets had pre-existing mutations that expanded (7%) or contracted (9%). Complex patterns were also observed (lower four panels) in which some subclones persisted simultaneously with the acquisition or loss of signaling mutations in other clones. (Adapted from Figure 3A and Figure 5D from Menssen AJ. Convergent clonal evolution of signaling gene mutations is a hallmark of myelodysplastic syndrome progression. Blood Cancer Discov. 2022;3[4]:330-345.)
Single-cell sequencing of a secondary acute myeloid leukemia (sAML) case from the study of Dr. Menssen and colleagues illustrates how the dynamic clonal architecture can be inferred by the subclonal mutation patterns. A) An initial clone carrying SRSF2 (spliceosome) and ASXL1 (epigenetic) mutations subsequently acquired a RUNX1 mutation (transcription) and then a PTPN11 mutation (signaling) as the final dominant clone in the sAML; a smaller subclone instead acquired NRAS, another signaling mutation. B) Complex clonal evolution pathways of signaling gene mutations in the progression from MDS to sAML. While a large subset of patients (26%) acquired a new signaling gene mutation upon development of sAML, significant subsets had pre-existing mutations that expanded (7%) or contracted (9%). Complex patterns were also observed (lower four panels) in which some subclones persisted simultaneously with the acquisition or loss of signaling mutations in other clones. (Adapted from Figure 3A and Figure 5D from Menssen AJ. Convergent clonal evolution of signaling gene mutations is a hallmark of myelodysplastic syndrome progression. Blood Cancer Discov. 2022;3[4]:330-345.)
In Brief
Single-cell sequencing of a secondary acute myeloid leukemia (sAML) case from the study of Dr. Menssen and colleagues illustrates how the dynamic clonal architecture can be inferred by the subclonal mutation patterns. A) An initial clone carrying SRSF2 (spliceosome) and ASXL1 (epigenetic) mutations subsequently acquired a RUNX1 mutation (transcription) and then a PTPN11 mutation (signaling) as the final dominant clone in the sAML; a smaller subclone instead acquired NRAS, another signaling mutation. Our current NGS methods have limited sensitivity to detect these clinically relevant low-level mutations, and more sensitive NGS technologies such as digital droplet and error-corrected PCR could be useful in the clinical arena. Second, they indicate that the presence of signaling pathway genes in low-risk MDS identifies patients who may have an unexpectedly high incidence of AML progression. While these results warrant validation in prospective studies, they suggest that low-risk MDS patients with signaling mutations (even if detected at an exceptionally low variant-allele fraction) could be monitored more closely for evidence for progression or even be considered for early intervention with treatments such as stem-cell transplantation. Finally, they reveal the complex and dynamic clonal architecture of dysplastic hematopoiesis; the MDS bone marrow is a “rough and tumble” place where different subclones compete, some expanding and others withering away over time. Ultimately, in some patients, a clone bearing a particular portfolio of mutations wins out and expands, giving rise to the devastating consequence of sAML.
While many current MDS and AML therapeutic strategies target specific mutations, the subclonal diversity of MDS revealed by the work of Dr. Menssen and colleagues and others suggests an alternate approach: trying to modify underlying factors that foster the expansion of deleterious clones. These factors could include inflammatory cytokines, altered gene expression in the malignant clone, or other elements in the bone marrow microenvironment.6 Although still beyond our reach, a valuable goal would be to develop therapies that discourage the outgrowth of small clones destined to grow into aggressive acute leukemia.
Competing Interests
Dr. Hasserjian indicated no relevant conflicts of interest.
