Abstract
Background: Pancreatic cancer (PC) is one of the most malignant cancers and its 5-year survival rate remains poor. Although immunotherapy has achieved certain therapeutic efficacy in some clinical trials, such treatment still shows low responsiveness and overall remission rate. Therefore, it is urgently necessary to dissect the tumor microenvironment and optimize the immunotherapeutic strategies against this malignancy.Methods: Using the multi-color immunohistochemistry, we investigated the expressions of B7-H3, B7-H4, HHLA2, CD8, and CD68 in 63 cases of PC tissues with the tissue microarray. Moreover, we analyzed immunolocalization features, prognostic values of this immune contexture, and clinical associations.Results: The expressions of B7-H3, B7-H4, and HHLA2 could be detected in cytokeratin (CK)+ tumor cells, CD68+macrophages, and even stromal cells. Higher expression of B7-H3 in tumor cells could predict a better survival of the PC patients. A positive correlation was found between the expressions of B7-H3 and HHLA2 in tumor cells, while there was a negative correlation between the expressions of B7-H4 and HHLA2 in tumor cells. A positive correlation was found between the expressions of B7-H3 and B7-H4 or HHLA2 in tumor-associated macrophages (TAMs), but not B7-H4 and HHLA2. Tumor-infiltrating CD8+T cells in combination with CD68+TAMs could serve as an important predictor for the postoperative prognosis of PC patients. Higher expression of B7-H3, or HHLA2 in CD68+TAMs could serve as an important predictor for poorer prognosis of PC patients. Patients with B7-H3lowB7-H4low, B7-H3lowHHLA2low, or B7-H4lowHHLA2low on CD68+TAMs could have a better postoperative prognosis compared with the other sub-populations in the combinational analysis.Conclusions: Taken together, our study indicated variable expressions and prognostic values of B7-H3, B7-H4, and HHLA2, in human PC tissues, and distinctively demonstrated that these immunosuppressive co-stimulators expressed by CD68+TAMs could be used as important bio-markers for the prognostic prediction of PC patients. Moreover, these results supported that the evaluation of these markers could be used as essential candidate targets for immunotherapy against PC.