Monocytic myeloid-derived suppressor cells in advanced melanoma patients

Abstract

The percentage of circulating CD14⁺CD11b⁺HLA-DR^−/low myeloid-derived suppressor cells (MDSCs) inversely correlates with survival among advanced melanoma patients. High levels of MDSCs are associated with the absence of T lymphocytes specific for melanoma-derived antigens, implying a causal and clinically relevant interaction between these cell subsets. MDSCs might therefore represent prognostic markers as well as targets for the development of novel therapeutic interventions against melanoma.

T cells are important effectors of the adaptive immune response. Indeed, they can destroy or maintain under control virus-infected as well as malignant cells upon recognition of specific peptides presented in complex with MHC molecules. Since the discovery of tumor-associated antigens (TAAs) more than 25 y ago, many groups worldwide have developed approaches to combat cancer through the activation of specific T-cell responses. Effective strategies of this type include adoptive T-cell transfer, a direct means to provide high amounts of tumor-specific T cells, as well as less direct manipulations such as the administration of immune checkpoint blockers or high-dose interleukin-2. Immunotherapeutic regimens that manipulate the T-cell response in melanoma patients can result in long-term tumor regression, yet clinical benefits are generally limited to a rather small subset of patients.

Fascinated by the favorable clinical course of some patients with unresectable advanced melanoma treated with cytotoxic T lymphocyte-associated protein 4 (CTLA4)-targeting antibodies at our department, we aimed to identify factors associated with potentially curative outcomes in long-term survivors that would set them apart from the majority of patients who do not experience such response. Initially, we identified melanoma antigen-specific T cells in the peripheral blood of most long-term survivors but few of the patients experiencing the usual clinical course (short-term survivors).¹ An analysis of the prognostic relevance of this finding in unresectable stage IV melanoma patients revealed that T cells, in particular those responding to NY-ESO-1- or Melan-A-derived antigens, are strong independent predictors of survival, even more robust than the M category of the AJCC staging system. Thus, patients with T cells responding to more than one of the 4 different melanoma-associated antigens that we tested had a better clinical outcome than individuals whose T cells responded to one or none.¹

In a subsequent study, we focused on the prognostic impact of cells that are capable of downregulating T-cell responses, notably regulatory T cells (Tregs) and myeloid-derived suppressor cells (MDSCs). Circulating CD4⁺CD25⁺FOXP3⁺ Tregs had no impact on prognosis but a high frequency of CD14⁺CD11b⁺HLA-DR^−/low MDSCs was strongly associated with survival in 133 late-stage melanoma patients bearing distant metastasis. The median survival time of patients with a low frequency of circulating MDSCs was 13 mo vs. 8 mo for other.² In this study, we confirmed the strong impact of melanoma-specific T cells and compared the relative prognostic effect of MDSCs, T cells and the M category by multivariate analyses. Despite the strong association with survival in univariate analyses, the circulating levels of MDSCs did not add independent prognostic information in Cox regression models. The reason for this unexpected finding is probably the strong correlation between the presence of melanoma-specific T cells and low MDSC levels implying a causal interaction between these cell types.² We cannot say whether the amount of functional T cells detected in our experiments reflects the in vivo levels of antigen-specific T cells or is significantly influenced by the amounts of MDSCs present during the 12 d of re-stimulation in vitro (Fig. 1). Further experiments are ongoing to clarify the impact of MDSCs on the detection of functional, antigen-specific T cells during in vitro expansion. We expect to observe a substantial influence of MDSCs in this setting. If this indeed turns out to be the case, the effects of MDSCs will have to be evaluated in assays commonly employed for the quantification of antigen-specific T cells after re-stimulation in numerous settings (e.g., ELISPOT assays).

Figure 1. Inverse correlation between antigen-specific T cells and myeloid-derived suppressor cells. We propose 2 hypotheses to explain the inverse correlation between antigen-specific T cells and myeloid-derived suppressor cells that we observed, both taking into account the fact that, in our assay, memory T cells are expanded to detectable levels in vitro. (A) Hypothesis 1: the correlation between low MDSC levels and detectable amounts of antigen-specific T cells (upper part) and high MDSC levels and undetectable T cells (lower part) reflects the in vivo situation. In this setting, MDSCs do not suppress T-cell expansion in vitro. (B) Hypothesis 2: the inverse correlation between the levels of MDSCs and antigen-specific T cells emerges in vitro, during expansion, and does not reflect the in vivo setting. Antigen-specific T cells are expanded to detectable amounts (irrespective of their abundance in vivo) when MDSC levels are low (upper part) but not in the presence of elevated frequencies of MDSCs (lower part).

MDSCs are populations of immature cells of myeloid origin, including precursors of macrophages, granulocytes and dendritic cells (DCs) at different stages of differentiation. MDSCs can be grouped into 2 main populations, exhibiting either a polymorphonuclear (PMN) or monocytic (M) phenotype.³ However, the functional immunosuppressive activity of these cells as measured in rather crude in vitro assays is currently the only common denominator of PMN- and M-MDSCs. Factors released by melanoma cells induce phenotypic changes in monocytes, leading them to resemble M-MDSCs found in advanced melanoma patients.⁴ Our work has indeed focused on M-MDSCs, which were originally found to be expanded in the peripheral blood of melanoma patients.⁵ M-MDSCs not only accumulate in the periphery but are also present within neoplastic lesions.³ Previous studies have shown that the levels of circulating MDSCs correlate with grading, stage, tumor burden,⁶ or clinical outcome in patients with different types of cancer.^7,8 Walter et al.⁸ reported a strong association between CD14⁺HLA-DR^−/low MDSCs and survival in vaccinated renal cell carcinoma patients, while Filipazzi et al.⁹ analyzed stage II/III melanoma patients with no evidence of disease and found a trend for improved disease-free survival among patients bearing low amounts of such MDSCs. Together with the results of these studies, our findings provide a strong rationale for the development of strategies for MDCS inhibition or depletion for the treatment of melanoma patients.

Glossary

Abbreviations:

M

monocytic

MDSC

myeloid-derived suppressor cells

PMN

polymorphonuclear

Citation: Martens A, Zelba H, Garbe C, Pawelec G, Weide B. Monocytic myeloid-derived suppressor cells in advanced melanoma patients: Indirect impact on prognosis through inhibition of tumor-specific T-cell responses?. OncoImmunology 2014; 3:e27845; 10.4161/onci.27845