Introduction
Breast cancer is the most commonly diagnosed cancer and the leading cause of cancer-related deaths in females worldwide [
1]. The striking mortality rates are caused by metastasis to distant regions from the primary tumor, while recent advances in improving survival rates are attributed to early detection through screening and initiation of neoadjuvant therapy in patients [
2]. The developments of novel approaches to identify prognostic markers for patients who are at high risk of developing breast cancer are, therefore, of cardinal significance.
Human carcinomas induce an immune response in the tumor microenvironment (TME) [
3]. However, emerging evidence has established the role of different immunosuppressive cells, such as myeloid-derived suppressor cells (MDSC) and regulatory T cells (Treg), in cancer bearing hosts. MDSC are a heterogeneous population of myeloid progenitor and activated myeloid cells, halted at varying stages of maturation and differentiation that exhibit a potent immunosuppressive activity against T-cell responses [
4]. Myeloid cells generated in bone marrow differentiate into mature granulocytes, macrophages or dendritic cells. However, pathological conditions such as cancers result in accumulation of a morphological mixture of granulocytic and monocytic MDSC in TME and circulation through various factors produced by tumors or activated T cells [
5]. The heterogeneous nature of MDSC together with functional and phenotypical overlap with other myeloid populations has, therefore, made it challenging to identify these cells. The majority of studies on MDSC in breast cancer have been carried out on murine models by developing spontaneous tumors due to predisposing mutations leading to breast cancer development or through transplantation of tumors in xenografts [
6]. Limited data are available for the conclusive phenotypical profiling of myeloid cells in circulation and the TME of breast cancer patients and their correlation with clinical settings.
MDSC in humans are commonly defined as cells which express common myeloid markers CD33 and CD11b but lack the expression of HLA-DR, and are further divided into monocytic MDSC (M-MDSC) or granulocytic MDSC (G-MDSC) based on the expression of CD14 and CD15, respectively, while immature MDSC (IM-MDSC) or early-stage MDSC (e-MDSC) lack CD14 and CD15 expression [
7]. MDSC exert their suppressive role through increased production of suppressive factors such as Arginase 1 (ARG1), nitric oxide and reactive oxygen and/or reactive nitrogen species along with modulating the production of various cytokines [
8]. Several studies have shown the phenotypical and functional similarities between G-MDSC and neutrophils [
9]. The term G-MDSC has recently been revised to polymorphonuclear MDSC (PMN-MDSC) to differentiate between steady-state neutrophils and G-MDSC, which have fewer granules and increased ARG1 and CD11b expression [
7]. Immunosuppression by tumor-associated neutrophils (TAN) uses similar mechanisms as MDSC and elevated neutrophil to lymphocyte ratio (NLR) is considered as a poor prognostic factor in cancer patients [
10].
In this study, we investigated the phenotypes and levels of myeloid cells in circulation and tumors from primary breast cancer (PBC) patients, and compared their levels with peripheral blood from healthy donors and paired, adjacent non-tumor breast tissue, respectively. We found that the immune profile of the TME of breast cancer patients is not reflected in circulation; there was an expansion of granulocytic and immature myeloid cells in the tumors but not in the periphery. Furthermore, there was no association between levels of circulating myeloid cells and patients’ TNM stage or histological grade. This disparity in peripheral blood and tumors provides a better understanding of the role of myeloid cells in the TME of breast cancer patients, and therefore, offers new facets for the development of therapeutic modalities to target the expanded immunosuppressive populations in the TME.
Discussion
Studying the immune profile of the TME in solid cancers is an active area of research. Tumor-infiltrating lymphocytes (TIL), tumor-associated macrophages (TAM), Treg and MDSC are widely recognized as prognostic or predictive markers for cancer progression. Early studies on lymphocytic infiltration in the TME of breast cancer patients showed a close correlation with disease prognosis [
15] and were found to have beneficial effects on survival [
16]. Several studies have since established certain subsets of TIL, mainly CD8
+ T cells as a good prognostic factor in various human cancers [
3,
17]. In contrast, tumor infiltration and expansion in periphery of immunosuppressive cells has been shown to correlate with poor prognosis and tumor progression in various human malignancies [
18]. MDSC and Treg are recognized as key players in the negative regulation of immune responses. Expansion of MDSC has been reported in different human cancers including head and neck, colon, renal, prostate and melanomas [
8]. Almand et al. reported an expansion of immature myeloid cells in peripheral blood from patients with head and neck and colon cancers that decreased after removal of tumors from these patients [
19]. However, limited data are available on MDSC levels in circulation and matched tumor tissues from cancer patients and their relation with clinical settings. The heterogeneous nature of MDSC due to the varying stages of differentiation at which they were halted, makes it challenging to identify these cells. Furthermore, this gives rise to a morphological mixture of cells, which share many phenotypical and functional characteristics of other cellular populations of myeloid origin. In the present study, we investigated phenotype and levels of tumor-infiltrating and circulating myeloid cells in untreated patients with primary breast cancer.
Several studies have emphasized on the significance of sample handling and processing when monitoring MDSC levels in circulation and are in agreement over the adverse effects of cryopreservation and the delayed time points at which MDSC analysis was carried out following blood collection [
20‐
22]. Mandruzzato et al. suggested to perform analysis on fresh blood to prevent possible loss of some MDSC subsets mainly G-MDSC. Fresh blood analysis also minimizes attenuation of cell surface markers due to Ficoll grade separation [
23]. Therefore, we used fresh whole blood for all our analysis on levels of myeloid cells in circulation.
Previous work on MDSC in breast cancer has mainly been performed on murine models. MDSC in mice express CD11b and Gr-1 with monocytic or granulocytic subsets identified by expression of Ly6C and Ly6G, respectively [
5]. Accumulating evidence in murine models has shown the significance of MDSC in development and progression of breast cancer [
24]. However, these models do not fully reflect the expression levels, cellular functionality of proteins or genes found in human breast tissue [
6].
We have recently shown that N/G-MDSC are expanded in the peripheral blood and the TME of patients with colorectal cancer [
25]. However, we did not find similar expansion of myeloid cells in circulation of breast cancer patients. Previous studies on MDSC in breast cancer patients have reported an expansion in patients mainly with advanced metastatic disease [
26‐
28]. However, the lack of consistency of markers and criterion to identify MDSC in these studies and more importantly the clinical presentation of study populations might account for discrepancies in results. Diaz-Montero et al. [
26] and Solito et al. [
27] identified MDSC as Lin
−/lowHLA-DR
−CD33
+CD11b
+ cells in patients with advanced breast cancers, while Yu et al. identified MDSC as CD45
+CD13
+CD14
−CD15
− cells with suppressive activity, assessed through IDO expression and reported expansion that correlated with lymph node metastasis in breast cancer patients [
28]. Interestingly, they reported that IDO expression was significantly upregulated in tumor-infiltrating MDSC than in periphery, thereby suggesting immunosuppressive role of MDSC in tissue and not in circulation [
28]. Additionally, Bergenfelz et al. reported an expansion of CD14
+HLA-DR
−/low monocytic MDSC in circulation in patients with metastatic breast cancer and this expansion correlated with disease severity [
29]. Another study showed an expansion of CD33
+HLA-DR
−/lowCD15
+CD11b
+ MDSC in peripheral blood of breast cancer patients with high psychological stress compared to those with lower stress levels, thereby suggesting an association between stress and immune function in breast cancer patients [
30].
The main finding in this study is that myeloid cells in the peripheral blood of this breast cancer cohort do not differ when compared to healthy individuals, but when assessing the levels in tumor vs surrounding healthy tissue, the levels were significantly higher in the TME. We found a significant expansion of CD33
+CD11b
+HLA-DR
−/low myeloid cells in the TME of breast cancer patients. These cells were mainly granulocytic, which include neutrophils and G-MDSC, along with IM-MDSC. Expansion of neutrophils results in suppression of cytolytic activity of immune cells and high NLR is associated with poor prognosis and reduced overall survival in various human malignancies [
31]. We also report an expansion of APC of myeloid origin in the TME of PBC patients, based on expression of HLA-DR, which is expressed only on professional APC. APC of myeloid lineage include DC and monocytes. Infiltration of DC has been reported in various solid tumors and has shown to be associated with both good and worse prognosis [
32,
33]. The association of tumor-infiltrating DC with worse prognosis in some cancers can be attributed to reduced antigen presentation of DC [
34]. Studies have shown tumor-induced functional deficiency of DC in breast cancer patients and reduced antigen-presenting functions in expanded Lin
−HLA-DR
+ cells in peripheral blood of cancer patients [
35]. Sathhaporn et al. showed defective DC function in peripheral blood from breast cancer patients with decreased IL-12 production, which could assist in tumor progression [
36], and Kitchler-Lakomy et al. showed reduced functional activity of DC in peripheral blood of breast cancer patients, which also exhibited immature morphology [
37]. Therefore, the expanded HLA-DR
+ cells in TME in this study require further functional investigation. Furthermore, association between MDSC levels and administration of therapeutic modalities have also been reported. Diaz-Montero et al. reported a significant increase of G-MDSC in breast cancer patients who received adjuvant chemotherapy [
26]. MDSC have been proposed as predictive markers for patients’ survival in various diseases. Bailur et al. demonstrated a negative role of MDSC and Treg in the prognosis of breast cancer patients by investigating the association between MDSC and CD8
+ cells in older untreated breast cancer patients [
38].
Breast cancer staging is widely accepted as a useful tool to estimate disease prognosis. Around 5–12% of breast cancer patients presenting with stage I or II tumors die within 10 years of diagnosis compared to over 60% of stage III and over 90% of stage IV patients [
39]. Diaz-Montero et al. reported a significant increase of MDSC in peripheral blood from patients with stage IV breast cancers which correlated with metastatic tumor burden [
26]. Solito et al. showed that the levels of immunosuppressive MDSC in stage IV advanced breast cancer patients correlated with circulating tumor cells and patients with higher levels had reduced overall survival compared to patients with lower levels [
27]. However, we did not find any correlation between circulating myeloid cells and patient staging, arguably due to the fact that patients in our study cohort, in contrast to these studies, presented with initial stages of cancer; none of the patients in our study group presented with distant metastasis. Histologic grading in breast cancers evaluates tubule formation, nuclear pleomorphism and mitotic count [
40]. Therefore, tumor histological grades are considered as stage-independent prognostic coefficients that reflect the metastatic potential of the tumor. We divided our cohort into two groups based on histological grades and compared circulating myeloid cell levels between them but did not find any difference between the groups.
Several proinflammatory molecules secreted by tumor cells are responsible for the recruitment of MDSC in the TME where they inhibit immune responses through various mechanisms, such as depletion of nutrients required for lymphocytes, inducing oxidative stress and activating Treg [
28]. ARG1 is involved in metabolism of
l-Arginine, required for T cell activation and reduces the expression of T cell receptor CD3ζ chain and impairs T-cell response [
41,
42]. In line with previous studies [
43], we investigated the immunosuppressive potential of myeloid cells by confirming the expression of ARG1 by circulating myeloid cells.
In conclusion, we showed that myeloid cells are expanded in the TME of breast cancer patients, and these cells comprise of immature and granulocytic myeloid cells. Interestingly, we did not find any expansion of myeloid cells in peripheral blood from breast cancer patients. These findings are of great significance in the development of therapeutic agents to target the mechanisms employed by immunosuppressive cells in providing an immune-permissive environment for the progression of cancer.