ABSTRACT
BACKGROUND AND OBJECTIVE: A reliable noninvasive prediction tool for the screening, diagnosis, and/or staging of colorectal cancer (CRC) before surgery is critical for the choice of treatment and prognosis. METHODS: Patients admitted for initial treatment of CRC between January 1, 2015, and December 31, 2018, were retrieved and reviewed. Records of CD16+CD56+ natural killer (NK) cells were analyzed according to the stages of CRC. RESULTS: The number of qualified participants in the healthy, stage I, stage II, stage III, and stage IV CRC patients were 60, 66, 60, 70, and 68, respectively. There was a significant difference in circulating CD16+CD56+ NK cells between the healthy group and the CRC group (p < 0.01), as well as between the healthy group and stage III or IV CRC group (p < 0.01 and 0.001, respectively). The percentage of circulating CD16+CD56+ NK cells in lymphocytes was negatively correlated with the occurrence of CRC. When comparing the pool of stage I and II CRC cases with the pool of stage III and IV CRC cases using circulating CD16+CD56+ NK cells, the area under the Receiver Operating Characteristic curve was 0.878. Using an optimal cutoff value of 15.6%, the OR was 0.06 (0.03, 0.11), p < 0.001, sensitivity was 86.5%, specificity was 72.5%, positive predictive value was 74.2%, and negative predictive value was 85.5%. CONCLUSIONS: Circulating CD16+CD56+ NK cells can be used as a screening and diagnostic/staging tool for CRC.