Ohri et al. indicated that a higher M1/M2 ratio led to an increased 5-year survival rate in non-small cell lung cancer patients, but they also found that the altered M1/M2 ratio were resulted from the significantly increased infiltration of M1 TAMs and was therefore not the result of an imbalanced TAM polarization process. Later, Edin et al. indicated that it was the number of infiltrating M1 TAMs, rather than the change in M1/M2 ratio, that substantially influenced the prognosis of colorectal cancer patients. However, a study conducted by Zhang et al. found the opposite. The authors observed that an increased M1/M2 TAM ratio alone was sufficient to predict a better prognosis for lung cancer patients. Soon thereafter, Algars et al. corroborated the finding that it was the M1/M2 ratio itself, rather than altered densities of the infiltrating M1 or M2 TAMs, that predicted a reduced long-term incidence of cancer relapse or hepatic metastasis in patients with colorectal cancer. Therefore, our results validated the findings of Zhang et al. and Algars et al. We demonstrated that an altered 
M1/M2 ratio alone is an independent prognostic indicator for ovarian cancer patients, which also implies that the chemotactic effect on monocytes/macrophages is not necessary for some immunosuppressive factors, such as MUC2, to influence patient prognosis. Moreover, considering that the reduced M1/M2 ratio has the highest HR, this pathological factor might have played a major role in the adverse outcome of MUC2++/+++ ovarian cancer cases. It was previously reported that cancer-derived MUC2 could initiate intracellular signaling by binding to the macrophage scavenger receptor on the surface of infiltrating monocytes/macrophages, promoting the upregulation of intracellular COX-2 gene expression in these cells. In this study, we re-examined this hypothesis in 102 cancer specimens. We found that most of the TAMs, which were closely surrounded by MUC2+ cancer cells, exhibited upregulated COX-2 expression. Moreover, we noted that this phenomenon was particularly significant in the high MUC2 expression group. These findings indicate that the local MUC2 expression level can help to determine the intratumoral density of COX-2+ TAMs, which implies its clinical significance. It has been known that COX-2 overexpression is associated with the increased synthesis of PGE2 in macrophages. The release of PGE2, which is a proinflammatory factor, can induce increases in the expression of VEGF, MMP, multi-drug resistance 1 and B-cell lymphoma 2 in the surrounding cancer cells, resulting in an improved vascularization of the cancer tissue and a reduced rate of apoptosis as well as an enhanced rate of metastasis and drug resistance in the cancer cell population. Hence, the existence of PGE2-releasing TAMs is generally an unfavorable prognostic factor. Because PGE2 is easily Foretinib degraded in the immunohistochemical labeling MK-2206 2HCl Akt inhibitor process, we had to immunohistochemically examined the COX-2 expression status in TAMs instead and used the obtained data for completing a Kaplan-Meier analysis. The result indicated that the COX-2 overexpression of TAMs was indeed associated with poor prognosis in the enrolled patients, suggesting that MUC2 also impaired patient survival via altering the local density of COX-2+ TAMs. In a seminal study, Torroella-Kouri et al. demonstrated that PGE2 could downregulate the transcriptional activity of NF-��B in immature monocytes and macrophages, which in turn reduces the expression levels of key M1-phenotype genes, such as iNOS and IL12. Recently, Nakanishi et al. confirmed that the administration of celecoxib, a selective COX-2 inhibitor, induced M2-polarized TAMs to become M1-polarized TAMs.