Abstract
We evaluate whether the tumor immune infiltrate (TIL) could be used for prediction of responsiveness to preoperative chemoradiotherapy (PCRT) in rectal cancers. Using formalin-fixed paraffin-embedded slides of pretreatment biopsies, co-stain for CD4, CD8, CD274 (PD-L1), FOXP3, cytokeratin, and DAPI was performed with Opal multi staining kit (Perkin-Elmer, Waltham, MA). Multispectral imaging and digital analysis to visualize and quantify specific immune infiltrates were performed using the Vectra imaging system (Perkin-Elmer). The density (number of cells per mm2) and proportion of total TILs and specific cell types in the stroma were calculated by inForm™ 2.2.1 software (Perkin-Elmer). The density and proportion of total TILs and specific cell types in the stroma were calculated by inForm™ 2.2.1 software (Perkin-Elmer, Waltham, MA). Patients were classified as group with total regression (TR, n = 25) and group with residual disease (near total, moderate, and minimal regression, RD, n = 50). The mean density of T cell infiltration and CD274 (PD-L1)+ lymphocyte were significantly higher in TR (p = 0.005, p = 0.001). The proportion of CD4+ lymphocyte (p=0.042) and CD274 (PD-L1)+ lymphocyte (p = 0.002) were different between 2 groups. The TR group has lower CD4+ and higher CD274 (PD-L1)+ proportions than RD group. The ratio among CD4+, CD8+, CD274 (PD-L1)+, FOXP3+ T cell was different between groups. TR group showed lower CD4/ CD274 (PD-L1) (p = 0.007), CD8/ CD274 (PD-L1) (p = 0.02), and FOXP3/ CD274 (PD-L1) (p = 0.003) ratio than RD group. The determination of the immune infiltrate in biopsies before treatment could be a valuable information for the prediction of responsiveness to PCRT.