Breast cancer is the most commonly diagnosed non-skin malignancy among women in the United States (US) and is the second leading cause of cancer-related death in women (1). Among prognostic factors used in clinical practice to determine the type of treatment indicated for each patient, the presence of metastatic axillary lymph nodes has been shown to be the most valuable, followed by expression of hormonal receptors [estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2/neu)], tumor size, histological subtype, tumor grade, lymphovascular invasion and proliferative rate. Although at present, metastatic axillary node status is the single most important prognostic factor, 20% of patients with histologically negative lymph nodes suffer recurrences and die of their disease within 10 years (2). Despite the existence of several prognostic factors, it remains a clinical challenge to predict clinical outcome. For these reasons, research is ongoing to identify better or more refined tumor prognostic markers resulting in more effective treatment choices.

Additional prognostic factors have been recognized but have not yet been validated for their use in breast cancer clinical practice. Some of these include tumor DNA content, amplification of oncogenes, loss of heterozygosity (LOH) of tumor-suppressor genes, angiogenesis and expression of proteases. Moreover, new prognostic tests based on tumor gene expression profiles such as Oncotype Dx, MammaPrint, Theros, MapQuant Dx, PAM50 and Mammostrat have been recently developed and are increasingly used for the prediction of clinical outcome in breast cancer patients (3–6). An additional factor associated with poor prognosis in breast and other solid malignancies is the presence of macrophages in the tumor microenvironment (7–11). There is increasing evidence that macrophages are implicated in the progression of a variety of cancers (12–14). The pro-tumor effects of tumor-associated macrophages (TAMs) include mutation induction, promotion of invasion, extracellular matrix remodeling, angiogenesis, metastasis and immune suppression (15,16).

Due to the generalized use of mammography and early detection screening programs, smaller T1 breast tumors (less than 2 centimeters in size across their widest point), are the most frequently detected in the US. When compared to Caucasian women, African-American and Latina women exhibit health disparities in breast cancer (17,18). Even though their genetic backgrounds are different, African-American and Latina women exhibit a similar pattern of breast cancer pathogenesis, with a slightly lower incidence but an earlier onset, and a more aggressive disease with less favorable prognosis (19).

Given the growing importance of macrophages in tumor progression, we examined the prognostic relevance of these cells in a small cohort of small T1 tumors across three ethnicities i.e. Caucasian, Latina and African-American women. We analyzed 30 total cases from the Sylvester Comprehensive Cancer Center Tissue Bank Core Facility’s retrospective archival tumor bank. This is a tumor bank containing samples from breast cancer patients treated at Jackson Memorial Hospital and University of Miami’s Sylvester Comprehensive Cancer Center in Miami, Florida between 1978 and 1996. Only T1 tumors that were ER⁺ by charcoal biochemical analysis (20) were selected. The charcoal method of ER determination was the one used in clinical practice to examine ER expression when these cases were diagnosed. We aimed to determine the prevalence of macrophages (CD68⁺), PR, HER2/neu and to confirm ER expression by immunohistochemistry (IHC) in these T1 breast cancers, which were all previously classified as ER⁺ by a charcoal biochemical method. Further, we analyzed any ethnic differences in marker staining and correlation with prognosis. We sought to evaluate whether CD68 expression in these breast cancers was associated with tumor prognosis and to assess any ethnic discrepancies between cases of Caucasian, African-American and Latina women with a pilot study of a small sample size. Our results demonstrated that macrophages are not major prognostic factors when evaluating small T1 ER⁺ breast cancers, although significant differences in TAM numbers are observed across ethnicities. Importantly, the lack of PR expression is indicative of poor tumor prognosis in this group.

It has been reported that inflammatory cells in the breast tumor microenvironment, particularly macrophages, contribute to tumor progression and are associated with poor tumor prognosis (8,11). However, there is a lack of studies that focus on the prognostic relevance of macrophages in small tumors. Here we used an immune scoring detection system to identify the presence and to determine the density of macrophage infiltrates in small T1 breast cancers, in order to assess whether these inflammatory cells can be used as independent prognostic factors to predict overall survival and to aid in decisions regarding adjuvant therapy in small breast tumors. We examined the expression of the pan macrophage marker CD68 using IHC in 30 T1 early breast cancers from a retrospective tumor bank, which provided the advantage that patient outcome was known.

Significant improvements in preventive medical care in the US during the last decades, with early detection programs and mammographic screening, have resulted in the majority of breast cancers being detected as small T1 tumors. However, despite this effort, a significant percentage of these breast cancer patients with small tumors still succumb to the disease, reflecting the imperative need to develop new and more refined prognostic markers which may lead to a more precise characterization of these tumors and to more effective treatments. The current study is among one of the first designed to evaluate whether the presence and density of CD68⁺ macrophages is an independent prognostic marker in small T1 ER⁺ breast cancers across three different ethnic groups. We were particularly interested in examining whether there were any differences in macrophage distribution in breast tumors among women from different ethnicities. In particular African-American and Latina patients presented with very aggressive and poor prognosis breast cancers, as compared with Caucasian women.

Selection of breast cancer therapy is based on standard prognostic markers, such as infiltration of axillary lymph nodes, tumor size, tumor grade and expression of hormonal receptors, although more recently, gene microarray technologies have also started contributing to this decision (24–26). In the present study, we determined essential clinical and histopathological characteristics of the cases examined: primary tumor size, involvement of regional lymph nodes, expression of hormonal receptors and status of ERBB-2 protein (HER-2), and their possible inter-correlation. The initial study included tumor samples from 30 women (10 Caucasian, 10 African-American and 10 Latinas). Importantly, despite these being T1 small breast cancers, 45% of the cases had some degree of lymph node involvement (cases with N1 and N2 were included and may have confounded the result, as being the strongest clinical pathologic factor for prediction), 41% were grades II and III, and one case had a distant metastasis. However, since all these cases were still within our inclusion criteria, we did not exclude them.

Breast cancer has been historically perceived as one disease with varying histopathological features and responses to systemic treatment. In the 1970s, however, breast cancer began to be divided into two disease subsets on the basis of ER expression, in view of the distinct clinical characteristics these subgroups display (27). The most widely used technique to determine ER expression in breast tumor samples before the introduction of IHC was the Dextran charcoal assay (28). This technique was based on a multipoint saturation analysis where a fixed amount of the tumor cytosol was incubated with increasing concentrations of labeled hormone. Dextran-coated charcoal (DCC) was used to separate bound from free hormone. Yet, results obtained with this assay showed high interlaboratory variations and exhibited many inconsistencies. Although real-time PCR and cDNA microarray have been employed to determine ER status in tumors, IHC is the currently used standard method to determine ER status in clinical tumor samples. This method recognizes ER through the use of monoclonal antibodies and provides fast and highly sensitive diagnoses. Our results included tumor samples from 29 women, since out of the 30 initial ER⁺ T1 cases that were selected for our study; one case was determined to be ER⁻ by IHC, despite being previously diagnosed as ER⁺ by charcoal biochemical method.

Although the occurrence of high numbers of macrophages in the tumor stroma has been associated with poor tumor prognosis in breast cancer in general (8), our results in small T1 breast cancers suggest no significant correlation between expression or density of the macrophage antigen CD68 and a reduced patient survival time using both univariate or multivariate survival analysis. Notably, Mahmoud et al(29) in a large cohort of breast cancer patients found that the presence of CD68⁺ macrophages was significantly associated with poor patient survival using univariate survival analysis, but that this association was not significant using multivariate survival analysis. Importantly, macrophage detection by IHC depends on the sensitivity of the marker used to identify this cell type. In this pilot study we used the classical pan macrophage marker CD68, which has been employed in the majority of studies. However, we are currently analyzing a larger cohort of breast cancers using the macrophage marker CD163, originally defined as an anti-inflammatory M2 macrophage marker (30) but more recently considered a pan macrophage marker, which has shown a greater sensitivity in our hands (data not shown) and in those of others (31) than CD68. Nevertheless, our present results conclude that CD68⁺ macrophage expression in small breast cancers is not a significant predictor of overall survival.

Although CD68 staining was not significantly associated with survival, we showed that the number of macrophages was significantly different among ethnicities when macrophage presence was assessed as positive or negative only, i.e. more tumors from Caucasian women were CD68-negative than those of African-American and Latina women. This result provides evidence to conclude that there is an association between the presence/absence of CD68 within the three ethnic groups. The presence of macrophages in small T1 breast tumors from African-American and Latina patients could be considered a contributive factor, among many other genetic, life style and social factors, which might in part explain the highest aggressiveness of breast cancers in African-American and Latina women. Indeed, Brown et al(32) demonstrated that race/ethnicity is a risk factor for survival in breast cancer.

Valuable staining information might be lost when one considers collapsing the number of macrophages to a dichotomous category as presence or absence other than considering them as macrophages i.e. ≥1 macrophages/slide. In our study, when instead of considering the presence or absence of macrophages, their numbers in the tumors were taken into account (as tumors with none, few, high, very high or super high numbers of macrophages), very strong and super strong CD68 expression was noted in the vast majority of tumors from African-American and Latina patients; Caucasian women mainly showed tumors without or with few macrophages, with the exception of a patient with very strong expression and another one with strong expression. Therefore, high macrophage density was also associated with ethnicity in our study, and the decision on how to properly categorize the IHC staining results should be made with great caution.

Breast cancers that are negative for ER, PR, and HER2 [called triple-negative (TN)] are associated with high-grade histology, aggressive clinical behavior and poor survival, and are highly prevalent in African-Americans and Latinas (32–34). An important finding from our study is the fact that the absence of PR expression even within small ER⁺ tumors is highly associated with poor tumor prognosis. Thus, similar to previous reports in non-T1 cases (35), the absence of PR is an independent prognostic factor for recurrence and poor clinical survival in ER⁺ breast cancer patients. ER⁺/PR⁻ tumors are a distinct subset of breast cancers characterized by aggressive behavior and tamoxifen resistance, and despite being ER⁺, they have a poor prognosis and are classified as luminal B cancers (36). Interestingly, and in contrast to our small cohort of small T1 breast cancers which were all HER2/neu-negative, the majority of these ER⁺/PR⁻ tumors were also HER2/neu⁺. It is important to point out that factors that may have limited our Cox regression analysis in the present study were the sample size and the number of events i.e. deaths. The major weakness of this pilot study was the fact that the number of the T1 tumor cases was relatively small, thus limiting subgroup analysis. Given these initial findings showing that CD68⁺ macrophages are not pivotal in determining tumor prognosis in early T1 breast cancers, new studies are presently being conducted with a larger sample size including the same three ethnicities, and comprising tumors from different stages, using various pan macrophage markers and also M1 and M2 macrophage activation markers.