Enhancing prediction performance by add‐on combining circulating tumor cell count, cd45^neg epcam^neg cell count on colorectal cancer, advance, and metastasis

Conventional circulating tumor cell (CTC) enumeration could ignore the CTCs more rel-evant to cancer metastasis. Thus, negative selection CTC enumeration was proposed, by which in-formation on two cellular biomarkers (numbers of CTCs and CD45^neg EpCAM^neg cells) can be ob-tained. By combining this approach with the conventional biomarker carcinoembryonic antigen (CEA), this study aimed to explore whether any combination of these biomarkers could improve the predictive performance for colorectal cancer (CRC) or its status. In this work, these two cell populations in healthy donors and CRC patients were quantified. Results revealed that enumeration of these two cell populations was able to discriminate healthy donors from CRC patients, even patients with non‐advanced CRC. Moreover, the combination of the two cell populations showed improved performance (AUROC: 0.893) for CRC prediction over the use of only one population. Compared with CEA alone, the combination of the three biomarkers increased the performance (AUROC) for advanced CRC prediction from 0.643 to 0.727. Compared with that of CEA alone for metastatic CRC prediction, the AUROC was increased from 0.780 to 0.837 when the CTC count was included. Overall, this study demonstrated that the combination of these two cellular biomarkers with CEA improved the predictive performance for CRC and its status.