Macrophages are one of the components of bone marrow cells and play roles in innate immunity. Classically, they circulate in the blood and migrate into the inflammatory tissues such as infectious diseases in response to specific, endogenous immune signal [1,2]. Recently their alternative roles in anti-inflammation, cell clearance and tissue regenerations have also been established especially in inflammatory metabolic disorders including obesity and diabetes [1,2]. Therefore, macrophages have been categorized into two subtypes-M1 and M2-depending on their distinctive roles. M1 macrophages play proinflammatory functions in response to TH1 cytokines and M2 macrophages play anti-inflammatory and imunosuppressive activities in response to Th2 cytokines [3]. However, macrophages are not fixed to a single M1 or M2 subtype but suspected to rapidly differentiate and redifferentiate in various microenvironments.

Tumor-associated macrophages (TAMs), macrophages existing tumor microenvironment, were first described in the early 1980s [4,5]. Till now, TAMs have been found to play dual functions, both positively or negatively affect tumor growth through interactions with the microenvironment, and these actions are tissue specific [6,7]. Several clinical studies showed that high density TAMs were present in more advanced stages of cancer with a worse prognosis in breast [8,9], lung [10,11], and bladder cancer [12,13]. Although a high TAM density has been reported in poorly differentiated papillary thyroid carcinoma (PTC) or anaplastic thyroid cancer [14], the distribution and function of TAMs in the most common PTC have not been studied yet. In this study, we have demonstrated the existence of TAMs in PTC tissue with lymph node (LN) metastasis and also used expression distribution and morphological properties to investigate the relationship between TAMs and the frequency of extrathyroidal invasion in PTC.

This study demonstrated the presence of TAM in a canopy structure over thyroid tumor cells in primary tumor tissues from 36 patients with PTC with LN metastasis. In addition, the primary tumors were significantly larger in the group with high TAM density. More LNs had metastases with a higher TAM density, although the difference was not statistically significant. Based on these findings, TAM density is considered to be related to PTC stage.

Ryder et al. [14] demonstrated a higher TAM density in poorly differentiated PTC and anaplastic thyroid cancer than in well-differentiated PTC, by using CD68 and CD163 immunohistochemistry. Qing et al. [15] also reported a high TAM density in well-differentiated PTC, and the TAM density was significantly higher in tumors with TNM staging over stage III. The results from these studies demonstrated that a high TAM density is present in poorly differentiated PTC and is related to poor prognosis, which indicates TAMs may support PTC growth. However, the role of TAMs in thyroid cancer remains controversial. Fiumara et al. [16], using CD68 immunohistochemistry, found that TAM prevented metastasis in some patients. Of 121 patients with PTC, 15% had TAMs CD68⁺) that appeared to have a phagocytic function on cancer cells. These patients had significantly less blood vessel invasion and remote metastasis and largely accompanying invasions of lymphocytes and dendritic cells, than patients without an CD68⁺ cells. These results suggest that these cells may activate the host immune response to cancer cells. However, in our study, 56% of patients had TAMs without evidence of phagocytosis, and their clinical characteristics were similar to those without TAMs.

This study investigated TAM morphologies and evaluated related clinical characteristics in PTC tissue with relatively advanced LN metastasis. Interestingly, CD68⁺ TAM cells observed in this study appeared to be mostly cytoplasm forming a canopy structure that projected over cancer cells. These canopy structures formed over the interior and exterior papillary projections in PTC cells that normally form papillary structures. These morphological characteristics are a standard of the TAM definition and differ from the rounded amoeba shape of inflammatory response cells, which are generally referred to as M1 macrophages [17]. The morphological characteristics of TAMs support cancer cell growth by forming a structural network between cancer cells and the extracellular matrix and are considered to support endothelial cells in forming blood vessels [18,19]. Caillou et al. [20] described the morphological characteristics of TAMs and their adjacency between cells, and these morphological characteristics suggest that TAMs facilitate cancer cell growth and metastasis. In this study, however, the morphological characteristics of thyroid cancer tissue were maintained regardless of the presence of TAMs, hence we suspect that TAMs do not have a huge effect on cancer development [20].

Interestingly, despite of small number of patients, when we classified patients according to TAM density, the primary tumors were larger in the high TAM density group. Hence, we hypothesize that TAMs inhibit tumor growth in advanced PTC. These results differ from previous findings, where Qing et al. [15] compared the micropapillary cancer smaller than 1 cm and papillary cancer of typical size (1 cm or larger) and found no difference in TAM density. Apart from the etiological interpretation, TAM can be interpreted to be involved in cancer invasiveness after mobilization followed by the cancer growth. Actually, TAM has been related to metastasis in various cancer types [12].

The study has a limitation of not including patients without LN metastasis. This study on TAMs reported that TAMs are rare in PTC, and especially they have only been reported in poorly differentiated cancer or anaplastic thyroid cancer. Therefore, this study only included the subjects with LN metastasis in higher stages of papillary cancer. A further study with more subjects including patients without LN metastasis is warranted.

Studies on cancer treatments targeting TAM have increased since the first report that TAM facilitates cancer growth and metastasis and offers indicates a poor prognosis. Popular treatment strategies that have been investigated involve cytokines such as colony stimulating factor (CSF)-1 [21], which mobilizes TAM to the tumor; interleukin (IL)-10 [21] and IL-4 [22], which facilitate a cancer's affinity for TAM through macrophage differentiation into the M1 and M2 subtypes; and CCR2 [23,24], which facilitates TAM in cancer metastasis. Interestingly, a recent animal study was published that suggests that these strategies might be feasible in treating thyroid cancer [25]. Ryder et al. [25] demonstrated PTC inhibition in a BRAF-mutant, transformed mice model by inhibiting CSF-1 gene expression through genetic modification or use of a CSF-1/CSF-1 receptor signal pathway inhibitor.

Studies on TAM in thyroid cancer are underway, but clinical data and studies on the mechanisms of TAM action are lacking. Thyroid cancer has a relatively favorable prognosis because its growth rate is slow. Therefore, general cancer pathophysiology does not apply to the role of TAMs. Currently the role of TAMs is only known in relatively invasive breast, lung, and pancreatic cancers, but the results of this study indicate that TAMs also exist in PTC, which is generally well differentiated with a favorable prognosis. Their role in cancer development is exciting; especially, the in light of the animal study targeting TAM that found possible inhibition of PTC. Expectations are high for treating patients with thyroid cancer since the BRAF mutation rate is higher among the Korean population than other ethnic groups. Additional study on the relationship between the BRAF mutation and TAM in Korean patients is warranted.

This study was conducted with relatively few patients, but it demonstrated the presence of TAMs in clinically common PTC patients, and the presence of a high TAM density in larger primary cancers. However patients without LN metastasis were not included in the study or the comparative analysis. Another limitation arose from not evaluating the characteristics of TAMs in metastasized LNs. Also, there is a fundamental controversy over whether CD68, the TAM marker, is appropriate for the staining M2 macrophages, so the more specific CD 163 should be stained in future analyses.

As a result of immunohistochemistry by using anti-CD68 antibody in PTC tissue with LN metastasis, TAM was present in all tissues with an extended cytoplasm forming a canopy shape over cancer cells. The density varied from 5% to 70% according to the tissue type. In normal thyroid tissue from the same patients, CD68⁺ cells were rare. When tumors were classified by TAM density, the primary tumors were statistically significantly larger in the group with a high TAM density. Therefore, TAM is suggested to stimulate cancer growth in PTC tissue with LN metastasis.