The Tumor Infiltrating Leukocyte Cell Composition Are Significant Markers for Prognostics of Radiotherapy of Rectal Cancer as Revealed by Cell Type Deconvolution

Tumor tissues consist of various types of cells including cancer stem cells, base cells, tumor infiltrated immune cells, and even new types of cancer associated cells. The tumor infiltrating cells play important roles in cancer progression and prognosis. Using present web servers and standalone tools of cell type deconvolution, including Timer, CIBERSORT, xCell, we deconvoluted the immune cell composition of microarray gene expression profiles of rectal cancer tissues collected from the rectal cancer patients before radiotherapy operation. By comparing the cell type composition obtained by timer of the rectal tissues from the radiotherapy responsive patients with those of non-responsive patients, we found that the content of CD4+ cells has average content of 0.1378 versus 0.1071 with p=0.0215; the ratio of CD4+/CD8+ is averaged 0.7869 versus 0.5564 with p=0.0210; and T cells CD4 memory resting and Eosinophils are both significantly higher with p-value of 0.033 and 0.0206, respectively, in radiotherapy responsive rectal cancer patient than in non-responsive cancer patients. The content of CD8+ cells are lower in the responsive than in non-responsive patients, which is averaged 0.1798 versus 0.2104 with p=0.0239. Other significant different cell types include macrophages M1 and M2, adipocytes, plasma cells and preadipocytes. Patients can be classified into responsive and non-responsive by machine learning method based on the tumor infiltrating immune cell composition with accuracy of 65%. It will be investigated that if the results can be applied to our newly collected data of other type of cancers including lung and nasopharyngeal carcinoma cancers in the future.