Background
A malignant tumor is comprised of cancer cells and the tumor microenvironment. The tumor microenvironment consists of extracellular matrix, endothelial cells, fibroblasts and leukocytes, all of which influence the fate of cancer cells and hence the clinical outcome. Tumor associated macrophages (TAMs) have been shown to enhance tumor progression by promoting tumor invasion, migration and angiogenesis [
1]. TAMs, which are often abundantly present in malignant tumors, share many common features with the alternatively activated anti-inflammatory macrophages (M2) [
2].
Macrophages can be differentiated into either pro-inflammatory M1 macrophages or anti-inflammatory M2 macrophages. M1 macrophages activate type 1 helper T cells (Th1), have the ability to kill pathogens and are tumoricidal. M2 macrophages on the other hand are involved in wound healing where they downregulate the inflammatory reactions, promote angiogenesis, recruit fibroblasts and regulate connective tissue remodeling. Furthermore, M2 macrophages have a weak tumoricidal capability [
3].
CD163 is a scavenger receptor upregulated by macrophages in an anti-inflammatory environment [
4] and regarded as a highly specific monocyte/macrophage marker for M2 macrophages [
5‐
7].
In 2002 Bingle et al. reported that the majority of publications on TAMs in cancer (including breast cancer) reported a correlation between high TAM infiltration and poor patient outcome [
8‐
11]. All papers in the metaanalysis except one (which used CD31 and thymidine phosphorylase) used CD68 as a marker for TAMs [
8]. However CD68, unlike CD163, recognizes both M1 and M2 macrophages [
12]. To our knowledge CD163 has not been evaluated as a TAM marker in primary breast cancer. Therefore, the aim of this study was to analyze whether CD163 can be used as a marker for TAMs and to compare its prognostic value with the more frequently used pan-macrophage marker CD68 in a tissue microarray (TMA) with tumors from 144 breast cancer patients.
In melanoma, the presence of TAMs in the tumor stroma (TS) correlated with poor overall survival (OS) [
13], while contrasting data have been reported for colorectal cancer [
14]. We therefore aimed to examine the importance of localization of TAMs on tumor-promoting capabilities and whether localization in the TS or tumor nest (TN) is of prognostic relevance in breast cancer.
Here we show that infiltration of CD163+ and CD68+ macrophages into TS, but not TN, is of clinical relevance for breast cancer patients. This highlights the importance of analyzing the localization rather than merely the presence of TAMs as a prognostic marker. While the presence of CD163+ macrophages in TS was more strongly associated with less favorable clinicopathological features, CD68+ macrophages in TS was a significant independent risk factor for a reduced breast cancer specific survival.
Discussion
CD68 is a pan-macrophage marker frequently used as a marker for TAMs. However, CD68 recognizes both tumoricidal M1 and anti-inflammatory M2 macrophages. This may explain previous contradictory data showing that dense infiltration of TAMs was correlated to both good and poor patient outcome [
8]. CD163 is a highly specific monocyte/macrophage marker that has recently been suggested to be expressed primarily by the anti-inflammatory subtypes of macrophages [
4‐
7,
25]. In this study we aimed to compare CD163 with the more frequently used pan-macrophage marker CD68. In addition we analyzed if the localization of macrophages in primary breast cancer could be of clinical relevance.
In our patient cohort, infiltration of TAMs in the TN was not correlated with any clinicopathological features and did not relate to OS, BCSS or RFS. The involvement of TN located TAMs for patient outcome should not however be fully excluded. In melanoma, for instance CD68
+ TAMs in the TN positively correlated with both poor OS and RFS, however it did not fall out as an independent risk factor upon multivariate analysis [
13]. In endometrial cancer and gastric cancer, dense infiltration of CD68
+ cells in the TN positively correlated with fewer recurrences [
26,
27] and hence suggested a beneficial effect of TAMs in the TN.
In our study, dense infiltration of CD163
+ macrophages located in TS correlated with grade, tumor size, subtypes and receptor status. In line with previous papers reporting that the content of TAMs inversely correlates with ER expression in breast cancer [
28,
29], CD163
+ macrophages located in TS correlated with ER negativity. Furthermore, CD163
+ macrophages in TS correlated with triple-negative/basal-like breast cancer and inversely with luminal A breast cancer. It has previously been suggested that the amount of TS correlated with a worse prognosis in triple-negative/basal-like breast cancer [
30]. In our cohort we did not find a correlation between the amount of TS and breast cancer subtypes, but we found an inverse correlation between the amount of TS and TS-associated CD163
+ and CD68
+ TAMs, rather strengthening the relevance of our findings. Further investigation is needed to understand what particular factors regulate the recruitment and activation of TAMs in the different tumor compartments.
Elkabets et al. recently reported that human tumors can attract GRN expressing hematopoietic cells, which promote malignancy by activating stromal tissue fibroblasts [
15]. These cells were located in close proximity to fibroblasts within the TS in mice. Notably, using the same breast cancer TMA as in this study, GRN expression was found to correlate with the same clinicopathological features as CD163
+ macrophages in TS [
15] as well as with the density of CD163
+ macrophages in TS. This suggests that the CD163
+ macrophages located in TS could represent GRN
+ hematopoietic cells.
Triple-negative/basal-like and luminal A breast cancers had different recruitment or differentiation patterns of CD163
+ and CD68
+ macrophages in TS. According to our data, triple-negative/basal-like breast cancers seem to harbor more CD163
+ than CD68
+ cells within the TS. One explanation for the difference between the presence of CD163
+ and CD68
+ macrophages in TS among triple-negative/basal-like breast cancer patients could be that CD68 is less expressed on mature M2 macrophages or monocytes. Another explanation could be that CD163 identifies another subset of cells. Although CD163 has been shown to be a specific monocyte/macrophage marker for IHC studies, flow cytometric analyses have shown that MDCs express CD163 [
24]. We can exclude that the majority of the CD163
+ cells are MDCs since the CD208
+ mature MDCs in this breast cancer cohort were only located in the peri-tumoral T cell rich areas. However, it cannot be fully excluded that there might be CD163
+ immature myeloid derived cells among the CD163
+ cells in TS. These CD163
+ immature myeloid derived cells could be myeloid derived suppressor cells which have been shown to enhance tumor progression by having an immunosuppressive effect on anti-tumor effectors [
31].
Although CD163+ macrophages in TS correlated with triple-negative/basal-like breast cancer it did not confer a prognostic value in this group. This could be due to the small size of the group and their worse overall prognosis, making it difficult to detect any significant differences in survival. However, it had a prognostic value for luminal A breast cancer patients.
Interestingly, CD68+ macrophages located in TS positively correlated with grade, tumor size and inversely correlated with luminal A breast cancer. It further correlated with poor OS and RFS in the univariate analysis. More importantly, high density of CD68+ macrophages in TS was an independent predictor of reduced BCSS. Even though CD163 and CD68 are expressed on the same cells and with strong correlations with each other, our data indicate that they also define immune subpopulations, which affect patient outcome differently and may therefore have distinct immune functions. In TS, the broader pan-macrophage marker CD68 had a better prognostic impact on BCSS compared to the anti-inflammatory CD163 marker, which is surprising since CD163 also seems to stain cells of myeloid origin other than TAMs. CD163+ cells were in particular abundant in triple-negative/basal-like breast cancer, which was a small patient group in our TMA. It would therefore be interesting to see if CD163 would have a prognostic value in a larger triple-negative/basal-like cohort.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
All authors made substantial contributions to the conception and design of the study, acquisition of data, analysis and interpretation of the data. Further they were involved in drafting the manuscript or revising it. All authors read and approved the final manuscript.