LAUSR

To the Editor, Serio et al. show a significant reduction of CD3CD56 regulatory T cells (TR3-56) in bone marrow (BM) of low-risk myelodysplastic subjects, as compared with the high-risk and the AML group; in addition, the BM frequency of mature granulocytes, a recognised marker of residual effective haematopoiesis, was observed to inversely correlate with TR3-56 in the MDS cohort. Such data are of great interest and confirm and extend, in an independent MDS cohort, the trend-increase of BM TR3-56 from very low/low risk to high/very high risk MDS and the inverse correlation with the cytotoxic T-cell (CTL) activity, likely fostering the escape of leukaemic blasts to immune-surveillance, by us recently described. Authors also observed BM TR3-56 frequency as negatively correlated with WT1 expression in AML, but not in MDS patients. Moreover, analysis of TR3-56 frequency after treatments in MDS patients, showed a persistent increase of this cell subset, regardless of therapy. Serio et al. conclude that BM TR3-56 frequency could be a flow-cytometry marker for MDS diagnosis, since consistently increased in high-risk MDS and AML, highlighting the role for this cell subset in promoting expansion of the dysplastic precursor/s with consequent ineffective myelopoiesis. Serio et al. data support the etiopathogenetic role of TR3-56 in MDS and the possible use of this population as diagnostic/prognostic marker of the disease. We previously suggested that type 1 diabetes (T1D) progression is associated with the loss of TR3-56-dependent control of CTL effector functions and, as commented by Serio et al., we highlighted the trend-increase of BM TR3-56 cells, with concomitant reduction of cytotoxic T-cell activity in MDS. The cytofluorimetric approach allows, in MDS, the evaluation of immune profile, in BM, the microsite in which deranged precursor/s maturation takes place; thus, the possibility that the analysis of TR3-56 might contribute to improve homogeneity of the diagnostic framing for MDS patients in whom an immune-mediated etiopathogenesis is conceivable, has to be also considered. Deranged activation and clonal expansion of the BM CTL represent a key element in the MDS pathogenesis. In very-low/lowrisk MDS patients, activated CTL and a pro-inflammatory environment contribute to impairment of polyclonal haematopoiesis, fostering the selection of dysplastic clones and their escape from immune control. Conversely, an immunosuppressive environment could disable CTL functions and favour the expansion/progression of dysplastic clones in the advanced stages of MDS. In very-low/lowrisk MDS subjects, the inverse correlation between BM TR3-56 amount and BM CTL activation and expansion suggest the possible participation of defective control of CTL effectors by the TR3-56 subset in immune-mediated mechanisms involved in the emergence of dysplastic clones, as proposed for the Treg subset. Mature granulocytes are significantly reduced in high-risk compared to low-risk MDS. Thus, the observation that the BM mature granulocytes frequency is inversely correlated with BM TR3-56, 1 intriguingly draws attention to the role of this regulatory subset in the etiopathogenesis and diagnosis of MDS.