Epigenetic Insights Into the Most Common Leukemia in Older Adults Revealed

Myelodysplastic syndromes can progress to secondary acute myeloid leukemia, a particularly aggressive and difficult-to-treat form of the disease. Understanding why some malignant cells survive treatment and ultimately drive this transformation is one of the major challenges in anticipating disease progression. A study published in the journal Leukemia has reconstructed the evolution of these malignant populations cell by cell and revealed that epigenetic changes play a key role in this process.

The research was led by Dr. Manel Esteller, head of the Cancer Epigenetics Group at the Sant Pau Research Institute (IR Sant Pau), ICREA research professor, and professor of Genetics at the University of Barcelona. The study combines analyses of DNA methylation, gene expression, mutational profiles, and cell-surface protein expression in individual cells to examine how myelodysplastic syndromes evolve before and after treatment and ultimately progress to secondary acute myeloid leukemia.

“Until now, the transformation of myelodysplastic syndromes into leukemia had been studied primarily from a genetic perspective, focusing on the accumulation of mutations. Our results indicate that epigenetic evolution also plays a key role and that different populations of malignant cells can follow distinct trajectories that ultimately lead to a more aggressive leukemia,” explains Dr. Manel Esteller.

A Disease That Can Progress to Aggressive Leukemia

Myelodysplastic syndromes are a group of blood cancers in which the bone marrow produces abnormal or insufficient blood cells. They are more common in older adults and, in some patients, can progress to secondary acute myeloid leukemia, which is often associated with resistance to conventional treatments and shorter survival. This transformation can occur even after a patient has received treatment and initially experienced clinical improvement.

One of the treatments used is azacitidine, a hypomethylating agent that modifies abnormal DNA methylation patterns. Methylation is an epigenetic mechanism that regulates gene activity without altering the DNA sequence and can determine which genes remain active or inactive in a cell. Although some patients initially respond to azacitidine, the disease can return or subsequently progress to leukemia, suggesting that some cell populations may survive treatment and retain the ability to drive disease progression.

Tracking the Disease Cell by Cell

The authors analyzed 15 bone marrow samples from five patients whose myelodysplastic neoplasms ultimately progressed to acute myeloid leukemia after treatment with azacitidine. Three stages of disease progression were studied for each patient: diagnosis, the period following treatment, and transformation into acute leukemia. Three patients had achieved a complete clinical response after treatment with azacitidine, while two were classified as nonresponders. Cells from healthy bone marrow were also analyzed as a reference.

The research focused on hematopoietic stem and progenitor cells with characteristics of malignant stem cells. These populations are particularly relevant because they can sustain the disease, survive treatment, and give rise to new tumor clones. Analyzing them individually makes it possible to detect differences between cells that would remain hidden in conventional studies conducted on heterogeneous samples containing thousands of cells.

“Conventional analyses provide an average value across thousands of cells and can conceal potentially relevant minority populations. Working at single-cell resolution allows us to see that the disease is not composed of a homogeneous population but of multiple subpopulations that evolve differently,” says Dr. Ignacio Campillo-Marcos, one of the study’s co-first authors.

Cells With Treatment-Resistant Epigenetic Profiles

The results showed an overall trend toward increasing alterations in methylation patterns from healthy bone marrow through transformation into acute myeloid leukemia. Overall, cells in the leukemic phase had a higher rate of epigenetic alterations than those observed when the myelodysplastic neoplasm was diagnosed. This evolution indicates that disease progression is accompanied by a greater diversity of epigenetic states and a progressive loss of stability in methylation patterns.

The researchers also identified differences between patients who responded to treatment and those who did not. At diagnosis, future responders did not have a significantly higher rate of epigenetic alterations than that observed in healthy bone marrow, whereas nonresponders already showed higher levels. Although the small number of patients means that this finding cannot be regarded as a clinical biomarker, the data raise the possibility that the initial degree of epigenetic alteration could eventually help predict treatment response.

“These results show that DNA methylation provides us with a layer of information that differs from genetics. What matters is not only which mutations the cells carry, but also how their genes are regulated and how much epigenetic diversity exists among them,” says Dr. Campillo-Marcos.

Treatment Alters the Epigenetic Landscape

After treatment with azacitidine, the samples showed a higher rate of epigenetic alterations and greater heterogeneity among cells than at diagnosis. This increase was observed both in patients who initially responded to treatment and in those who did not, indicating that treatment-associated epigenetic diversification cannot, by itself, be interpreted as a sign of clinical efficacy.

In responders, azacitidine substantially reduced the number of malignant stem cells and mutation-bearing clones. However, the cells that remained displayed a high degree of diversity in their epigenetic patterns. This observation suggests that treatment may eliminate or reduce certain cell populations, while others survive and retain epigenetic states that could promote their persistence and progression to acute leukemia.

“Treatment does not act on a homogeneous population of malignant cells, but on an ecosystem of clones with different epigenetic states. Some cells may have profiles that give them an evolutionary advantage, enable them to better withstand therapeutic pressure, and facilitate their subsequent expansion,” explains Dr. Esteller.

The authors emphasize that these results do not indicate that azacitidine causes disease progression. The drug exerts pressure on the different cell populations: it eliminates or reduces some of them, while others may survive and ultimately contribute to relapse or leukemic transformation. The study also identified differences in methylation of the NPHP4 gene between responders and nonresponders, consistent with previous research linking its methylation status to the response to hypomethylating agents. However, this potential association will need to be validated in larger groups.

Clones That Foreshadow Future Leukemia

Using the methylation data, the researchers reconstructed epiphylogenetic trees representing the relationships among individual cells throughout disease progression. The length and distribution of the branches reflect accumulated differences in methylation patterns and make it possible to identify which cells share similar epigenetic states at each stage.

In the two patients who did not respond to treatment, cells from samples obtained after treatment with azacitidine were found on the same branches as cells from the subsequent leukemia. This means that before the transformation became clinically detectable, some populations already shared methylation patterns with cells later observed in the leukemic phase. The results are consistent with the possibility that these cells are selected during treatment and contribute to treatment resistance and disease progression, although the study does not directly demonstrate that they subsequently give rise to leukemia.

“This finding is important because it indicates that leukemic transformation does not occur suddenly. Cell populations may evolve during treatment and promote progression long before the disease becomes clinically apparent,” Dr. Esteller emphasizes.

Changes in Gene Activity

The researchers also analyzed the gene expression of 104,919 cells and identified 35 distinct cell populations. In responders, azacitidine was associated with a reduction in malignant stem cells located within hematopoietic stem cell and myeloid progenitor compartments. This decrease was not observed in patients who did not respond to treatment, although the authors caution that the effect recorded in the responder group was driven primarily by one of the three patients.

The study also identified 27 genes in which methylation was significantly associated with expression levels. The leading candidates included STAT5B, a gene that plays a significant role in the formation and differentiation of myeloid cells. This association shows that some methylation changes detected are accompanied by changes in gene function and do not merely represent descriptive differences between cells.

In responders, treatment was associated with the activation of cell-cycle control and DNA-repair mechanisms in the cellular compartments containing malignant stem cells, probably in connection with the accumulation of DNA damage caused by azacitidine. At the same time, the activity of signaling pathways associated with FLT3 and MAPK decreased in these same cell populations. These pathways are frequently linked to unfavorable outcomes in acute myeloid leukemia. These cells also showed lower expression of genes associated with two signatures linked, respectively, to acute myeloid leukemia stem or progenitor cells and to a poorer prognosis in myeloid neoplasms.

Genetics and Epigenetics Act Together

To complete the analysis, the authors studied mutations in 53 genes associated with myeloid neoplasms and the expression of 42 cell-surface proteins in 17,849 cells. Alterations were identified in 12 genes involved in different biological processes: epigenetic and chromatin regulators such as EZH2, TET2, IDH1, and ASXL1; genes associated with transcription, such as RUNX1 and BCOR; genes involved in RNA processing, such as SF3B3 and U2AF1; and genes linked to cell signaling, such as NRAS, PTPN11, FLT3, and NPM1.

In 80% of cases, the earliest genetic events affected epigenetic regulators. Mutations associated with transcription and RNA processing appeared later, while the latest alterations tended to impact signaling pathways, particularly NRAS and FLT3. However, the researchers found no direct relationship between the type of mutated epigenetic regulator and the degree of DNA methylation heterogeneity, indicating that epigenetic diversity cannot be explained solely by the mutations present.

In patients who responded to azacitidine, the number of mutated clones decreased markedly after treatment. During the transformation into leukemia, however, the researchers observed the persistence of some clones, the expansion of populations that were initially in the minority, or the emergence of new mutated clones. One patient acquired an FLT3 mutation during progression. According to the researchers, this alteration may have given the malignant cells an advantage that enabled them to overcome the constraints imposed by treatment and regain their capacity for clonal expansion.

“Genetics and epigenetics do not evolve independently. Mutations can provide certain cells with advantages, but their behavior also depends on their epigenetic state and the expression programs active at any given time,” explains Dr. Esteller.

Toward Medicine That Can Anticipate Disease Progression

The results support a model in which multiple populations of malignant stem cells evolve in parallel during the progression of myelodysplastic neoplasms. Rather than following a single linear trajectory, these populations compete, disappear, persist, or expand depending on their genetic and epigenetic characteristics and the pressure exerted by treatment. Transformation into leukemia would therefore result from a dynamic ecosystem in which different clones coevolve and may acquire advantages at different times.

This perspective could eventually make it possible to identify residual populations at greater risk of driving leukemic transformation and develop treatments targeting epigenetic states shared by genetically distinct cells. “Now that we have identified changes in the composition of the epigenome during disease progression, it will be important to study how to eliminate this small subgroup of cells with the potential for progression, transformation, and drug resistance,” concludes Dr. Esteller.

The authors identify the small number of patients analyzed as the study’s main limitation, meaning that the work should be considered exploratory. The results will need to be validated in larger, well-characterized clinical cohorts to determine their potential applications in diagnosis, prognostic prediction, and treatment selection. Further research will also help clarify how genetic and epigenetic alterations interact and whether the populations detected after treatment can be used to anticipate and prevent transformation into leukemia.

Reference Article:

Bueno-Costa A, Campillo-Marcos I, Casado-Peláez M, Yassine K, Noguera-Castells A, Zamora L, Xicoy B, Solé F, Mata C, Park J, Ganesan S, Della Porta MG, Ferrer G, Landau DA, Esteller M. Single-cell DNA methylation analysis uncovers epigenetic pathways in the transformation of MDS to AML. Leukemia 2026. https://doi.org/10.1038/s41375-026-03015-z.