Myelodysplastic syndromes (MDS) are blood disorders in which the bone marrow does not effectively produce healthy cells. Their progression can be unpredictable: some patients are at risk of developing a more severe form of leukemia, secondary acute myeloid leukemia (sAML). To try to predict this risk, physicians classify patients as “low” or “high risk” based on established prognostic systems. However, this classification does not always reflect the biological complexity of the disease: some patients considered low risk may still deteriorate rapidly.

A recent study published in Blood found that very early changes in gene regulation and activation can reveal a hidden subgroup of low-risk patients already predisposed to disease progression. By analysing both DNA organisation and gene activity, researchers highlighted the role of PU.1, a protein that coordinates the activity of many genes and appears to link immune response to disease progression.

The study, supported by the AIRC Foundation for Cancer Research,was led by Serena Ghisletti, co-corresponding author, from the European Institute of Oncology and the University of Milan and Matteo Giovanni Della Porta, Head of the Leukemia Unit at Humanitas Research Hospital and Professor at Humanitas University, with contributions from numerous researchers and clinicians in his group, together with the Laboratory of Clinical and Experimental Immunology led by Domenico Mavilio. Other Italian institutions, including Human Technopole and CNR, also participated.

“Our results show that low-risk patients are not all the same: some already display molecular features indicating a higher likelihood of deterioration,” explains Della Porta. “This allows us to envision more precise prognostic models that integrate molecular and epigenetic information.”

A “silent” subgroup at risk of progression

To understand why some patients worsen despite being classified as low risk, the researchers studied bone marrow stem cells (CD34⁺), from which all blood cells originate, in patients at different disease stages. By analysing both which genes are active and how DNA is organised, they identified distinct molecular patterns associated with disease progression.

This revealed a subgroup of patients formally classified as low risk but with genetic features similar to those seen in high-risk or sAML patients. These patients show increased activation of the immune system and inflammatory processes, with more active T lymphocytes and natural killer cells, as well as mutations in key genes that regulate how genetic information is “assembled,” such as SRSF2. Clinically, this translates to a higher susceptibility to infections and cardiovascular events, a greater risk of disease progression, and reduced overall survival.

“We discovered a mechanism through which epigenetic changes trigger immune and inflammatory responses, driven by the PU.1 factor,” adds Ghisletti. “This helps us understand how early molecular alterations can shape disease evolution.”

PU.1: the “engine” of progression

PU.1 is not only more abundant in diseased cells but also changes how it binds to DNA, activating genes that regulate immune and inflammatory responses.These changes occur even in the early stages of the disease in low-risk patients.

Laboratory experiments confirmed that PU.1 supports the disease: when its activity is blocked, MDS cells proliferate less and show reduced self-renewal, while the activity of immune and inflammatory genes decreases. In other words, PU.1 is not just a marker of progression—it actively drives it.

Towards more precise medicine

These findings suggest that early molecular changes could be used to predict which low-risk patients may deteriorate, offering a more detailed view of the disease and its evolution.

Identifying this high-risk subgroup could improve clinical management and guide targeted therapeutic strategies. PU.1, in particular, represents a promising target for future therapies: blocking its action could disrupt the mechanisms driving disease progression.

Although further studies are needed to translate these discoveries into clinical treatments, integrating epigenetic information into existing prognostic models represents an important step toward more personalised and precise medicine for MDS.