Characterization of unique pattern of immune cell profile in patients with nasopharyngeal carcinoma through flow cytometry and machine learning.

Abstract

In patients with nasopharyngeal carcinoma (NPC), the alteration of immune responses in peripheral blood remains unclear. In this study, we established an immune cell profile for patients with NPC and used flow cytometry and machine learning (ML) to identify the characteristics of this profile. After isolation of circulating leukocytes, the proportions of 104 immune cell subsets were compared between NPC group and the healthy control group (HC). Data obtained from the immune cell profile were subjected to ML training to differentiate between the immune cell profiles of the NPC and HC groups. We observed that subjects in the NPC group presented higher proportions of T cells, memory B cells, short-lived plasma cells, IgG-positive B cells, regulatory T cells, MHC II+ T cells, CTLA4+ T cells and PD-1+ T cells than subjects in the HC group, indicating weaker and compromised cellular and humoral immune responses. ML revealed that monocytes, PD-1+ CD4 T cells, memory B cells, CTLA4+ CD4 Treg cells and PD-1+ CD8 T cells were strongly contributed to the difference in immune cell profiles between the NPC and HC groups. This alteration can be fundamental in developing novel immunotherapies for NPC.