Granulocytic myeloid-derived suppressor cells in peripheral blood of patients with cutaneous melanoma

To the Editor-in-Chief,

The article published in this issue (p. XX) by Stanojevic et al. (1) describes a myeloid-derived suppressor cell (MDSC) subset that is increased in melanoma patients and shows further accumulation during disease progression. Importantly, these cells were identified and quantified in whole blood. The authors suggest that MDSC measurements in whole blood should be considered as a potential biomarker. Although a very interesting study, several weaknesses warrant further discussion.

We would like to point out that it is hard to distinguish conventional neutrophil granulocytes or eosinophils from granulocytic MDSC (grMDSC), especially in whole blood (2). The authors found grMDSC in whole blood to be CD10^–CD15⁺CD14^–HLA-DR^–/lowCD33^lowCD11b⁺CD45^lowCD16^low and lineage negative, which is a widely used panel for identifying grMDSC in PBMC after density centrifugation (2). Neutrophil granulocytes show the same marker expression but are also positive for CD16. However, in elderly patients reduced CD16 expression on neutrophil granulocytes is a common finding and can contribute to immunosenescence (3). In addition, it is known that neutrophil granulocytes can lose CD16 when they undergo apoptosis (4) and surface expression of CD16 is reduced upon activation (5). Apoptotic down-regulation might also explain the lower expression of CD33, CD11b, CD45 and CD10 on those cells considered grMDSC by Stanojevic et al. compared with regular neutrophils. As discussed by the authors, human eosinophils also share the phenotype attributed by the authors to grMDSC (1). However, distinguishing eosinophils and grMDSC on the basis of low or absent CD16 expression does not appear sufficient in the context of using grMDSC as a biomarker. In their study, the authors also compare grMDSC in whole blood and after density-gradient centrifugation. Nevertheless, activated and immature granulocytes can show altered buoyancy in density gradients as well, which can result in altered sedimentation properties (2). Finally, human MDSC are better defined by their functional property of suppressing T-cell functions than by a phenotypic description. Since such evidence is lacking, we strongly suggest isolating these cells considered to be grMDSC by Stanojevic et al. and assessing their functional properties.

We agree that a method able to identify grMDSC in whole-blood samples is needed to explore the value of these cells as a biomarker. The manuscript ‘A subpopulation that may correspond to granulocytic myeloid-derived suppressor cells reflects the clinical stage and progression of cutaneous melanoma’ provides preliminary evidence that such a group of cells corresponds with tumor progression and can be identified in whole-blood samples. However, it remains unclear if the approach suggested by Stanojevic et al. is suitable to identify grMDSC in whole blood or if the leukocyte population described might be consistent with immature, activated or apoptotic neutrophil or eosinophil granulocytes.