Abstract:

Objective To study the five EB virus antibody profiles in normal physical examination people and nasopharyngeal carcinoma（NPC）patients in Guangzhou，we try to establish a simple，efficient and practical serological screening method，which is suitable for NPC screening in the region. Methods A total of 204 serum samples of the patients with NPC were collected. A total of 2，829 healthy people′s serum samples were collected in Guangzhou cadre Health Management center from December 2019 to November 2020，and the male ⁃female ratio was close to 3∶1. The antibody profiles and the differences of five kinds of EB virus antibody expression between two groups were analyzed，a predictive model for NPC by serological indicators was established through Logistic regression analysis. Results The optimal cut⁃off values of VCA/IgA，NA1/IgA，Rta/IgG，EA/IgA and EA/IgG for the diagnosis of NPC were 1.19，0.31，0.33，0.62 and 0.83，respectively. The AUC was 0.948，0.916，0.951， 0.873 and 0.942，respectively. There was no significant difference in the positive rates of five antibodies between the disease group and the control group（P > 0.05）. The predictive coincidence rates of the two combined predictive models for NPC were 96.5% and 96.2% respectively，and the negative predictive values of both models were greater than 98%. Conclusion The combined detection model of VCA/IgA + NA1/IgA and NA1/IgA + Rta/IgG have good sensitivity and specificity in predicting NPC all. The accuracy of the prediction model in determining non⁃nasopha⁃ ryngeal carcinoma in physical examination population in this area is more than 98.8 %.

Key words:

nasopharyngeal carcinoma, EB virus antibody, prevalence prediction model, tumor screening