RÉSUMÉ
PURPOSE: Little is known about the clinical value of peripheral blood immune profiling. Here, we aimed to identify colorectal cancer (CRC)-related peripheral blood immune cells and develop liquid biopsy-based immune profiling models for CRC diagnosis. METHODS: Peripheral blood from 131 preoperative patients with CRC and 174 healthy controls was analyzed by flow cytometry and automated hematology. CRC-related immune factors were identified by comparing the mean values of immune cell percentages and counts. Subsequently, CRC diagnostic algorithms were constructed using binary logistic regression. RESULTS: Significant differences were observed in percentages and counts of white blood cells, lymphocytes, neutrophils, regulatory T cells, and myeloid-derived suppressor cells (MDSCs) of patients and controls. The neutrophil/lymphocyte and Th1/Th2 ratios were also significantly different. Likewise, the percentages and counts of peripheral blood programed death 1, cytotoxic T lymphocyte antigen 4, B-and T-lymphocyte attenuator, and lymphocyte activation gene-3 were higher in patients with CRC. The binary logistic regression model included 12 variables, age, CD3+%, NK%, CD4+CD279+%, CD4+CD25+%, CD4+CD152+%, CD3+CD366+%, CD3+CD272+%, CD3+CD223+%, CD158b−CD314+CD3−CD56+%, Th2%, and MDSCs cells/µL, for the prediction of cancer. Results of retrospective and prospective evaluation of the area under the curve, sensitivity, and specificity were 0.980 and 0.940, 91.53% and 85.80%, and 93.50% and 86.20%, respectively. CONCLUSION: Peripheral blood immune profiling may be valuable in evaluating the immunity of CRC patients. Our liquid biopsy-based immune diagnostic method and its algorithms may serve as a novel tool for CRC diagnosis. Future largescale studies are needed for better characterization of its diagnostic value and potential for clinical application.