Abstract Myelodysplastic syndromes (MDS) are a heterogeneous group of hematological malignancies with an increased risk of evolving into Acute Myeloid Leukemia (AML). Novel and powerful agents have recently entered the clinical stage, but the overall prognosis of MDS and AML patients is still poor. Cell-autonomous defects are pivotal in the development of MDS and AML. However, recent evidence also demonstrated that microenvironmental cell-extrinsic factors, particularly inflammatory signals in the bone marrow (BM), play a crucial role in MDS. In most tumors, the inflammatory mediator interferon (IFN)-γ is a favorable predictive marker, providing a downstream signal that results in enhanced anti-tumor immunity. We have revealed an unexpected and unwelcome tolerogenic role of IFN-γ in AML. We demonstrated that leukemic cells produced IFN-γ, allowing us to stratify them into IFN-γ-high and IFN-γ-low samples. Higher IFNγ levels remodeled the BM microenvironment, favoring the indoleamine 2,3-dioxygenase (IDO)1-dependent induction of immunosuppressive Tregs, which was associated with increased engraftment of leukemic cells in a mouse model. This study aims to explore the induction of Tregs as putative IFNγ-dependent mechanisms of tolerance and immune escape in MDS and their evolution in AML. MDS samples showed IFN-γ levels higher than those in IFN-γ-high and IFN-γ-low AML samples. We compare IFN production in all components of the BM ME between the analyzed samples and can distinguish three different types of ME. IFNγ-low AML samples are ‘cold’ with the lower levels of inflammation in all BM components; IFNγ-high AML samples are ‘hot’ with IFNγ mainly produced by blasts; MDS samples are the hottest with higher IFNγ in dysplastic cells as well as in other BM components, such as CD3+, CD8+ T cells, and NK cells. Interestingly, we found a positive correlation between IFNγ and IDO1 expression at both the mRNA and protein levels, suggesting an immune-tolerant microenvironment in MDS that is similar to AML. MDS samples had a higher percentage of total Tregs than healthy donors and IFNγ-low but lower than IFNγ-high AML samples. However, in MDS, BM Tregs varied with prognostic risk but were not correlated with IFN-γ levels, suggesting that a mechanism of immunosuppression different from that in AML operates in MDS. Our results showed that IFNγ production by MDS cells remodels the BM landscape toward IDO1 upregulation but not Treg induction. Overall, the study emphasizes the importance of understanding the complex interplay between immune factors and the microenvironment in MDS, which is emerging as a critical aspect in MDS risk stratification and progression in AML.
Ciciarello et al. (Mon,) studied this question.