Higher correlation between neutralizing antibodies and surrogate neutralizing or binding antibodies in COVID-19 patients than vaccine recipients

This study determined the seropositive rates and levels of antibodies to severe acute respiratory syndrome coronavirus-2 in 50 patients and 108 vaccinees using microneutralization test (MNT), surrogate virus neutralization test (sVNT), chemiluminescent microparticle immunoassay (CMIA), and electrochemiluminescence immunoassay (ECLIA). MNT, as the reference method, employed living clade S and Delta viruses to measure neutralizing (NT) antibodies, while sVNT employed wild type strain and Delta receptor-binding domains (RBD) as the test antigens to measure sVNT antibodies. CMIA and ECLIA employed only one version of RBD to measure the binding antibodies. Our study performed S gene sequencing of the test virus to exclude undesired mutants that might lead to changes in antibody levels in MNT assay. We showed that spike protein amino acid sequences of our Delta virus contained 13 amino acid changes, with 3 related to the reduced neutralization. The MNT assay showed a significant reduction in seropositive rates and antibody levels in the patients’ sera when the Delta variant replaced clade S as the test virus. In contrast, the seropositive rates determined by sVNT assay using wild type strain RBD and Delta RBD were non-significantly different, suggesting that sVNT assay could not identify the difference between the antigenicity of wild type RBD and Delta RBD. Furthermore, the correlation between the levels of NT and sVNT antibodies was moderate with the patients’ sera but modest with the post-vaccination sera. The seropositive rates in the patients, as determined by CMIA or ECLIA, were not different from the MNT assay using clade S, but not Delta, as the test virus. In all analyses, the correlations between the antibody levels measured by MNT and the other 3 assays were modest to moderate, with the r-values of 0.3500–0.7882.