Evaluation of seven recombinant VCA-IgA ELISA kits for the diagnosis of nasopharyngeal carcinoma in China: a case-control trial.

Abstract

Seven recombinant viral capsid antigen-IgA (VCA-IgA) ELISA kits are widely used in China, but their diagnostic effects have not been evaluated. In this study, we evaluated whether the diagnostic effects of these kits are similar to those of the standard kit (EUROIMMUN, Lübeck, Germany). A diagnostic case-control trial was conducted with 200 cases of nasopharyngeal carcinoma (NPC) and 200 controls from NPC-endemic areas in southern China. The areas under the curve (AUCs), the sensitivities and the specificities of testing kits were compared with those of the standard kit. The test-retest reliability of each kit was determined by intraclass correlation coefficient (ICC). Their diagnostic accuracy in combination with Epstein-Barr virus nuclear antigen 1-IgA (EBNA1-IgA) was also evaluated in logistic models. Three testing kits-BB, HA and KSB-showed diagnostic accuracy equal to that of the standard kit, with good performance in the AUCs (0.926-0.945), and no significant differences in sensitivity were found between early-stage and advanced-stage NPCs. ICCs exceeded 0.8. Three logistic regression models were built, and the AUCs of these models (0.961-0.977) were better than those of the individual VCA-IgA kits. All new models had diagnostic accuracy equal to that of the standard kit. New cut-off values of these three kits and their corresponding combinations for researchers to replicate and use in NPC early detection and screening in the future were provided. Three recombinant VCA-IgA kits-BB,HA and KSB-had diagnostic effects equal to those of the standard kit, and, in combination with EBNA1-IgA in logistic regression models, can be used in future screening for NPC.