Evaluation of a surrogate virus neutralization test for high-throughput serosurveillance of SARS-CoV-2

Abstract

High-throughput serological tests that can detect neutralizing antibodies against SARS-CoV-2 are desirable for serosurveillance and vaccine efficacy evaluation. Although the conventional neutralization test (cVNT) remains the gold standard to confirm the presence of neutralizing antibodies in sera, the test is too labour-intensive for massive screening programs and less reproducible as live virus and cell culture is involved. Here, we performed an independent evaluation of a commercially available surrogate virus neutralization test (sVNT, GenScript cPass™) that can be done without biosafety level 3 containment in less than 2 h. When using the cVNT and a Luminex multiplex immunoassay (MIA) as reference, the sVNT obtained a sensitivity of 94 % (CI 90-96 %) on a panel of 317 immune sera that were obtained from hospitalized and mild COVID-19 cases from Belgium and a sensitivity of 88 % (CI 81-93 %) on a panel of 184 healthcare workers from the Democratic Republic of Congo. We also found strong antibody titer correlations (r_s>0.8) among the different techniques used. In conclusion, our evaluation suggests that the sVNT could be a powerful tool to monitor/detect neutralising antibodies in cohort and population studies. The technique could be especially useful for vaccine evaluation studies in sub-Saharan Africa where the basic infrastructure to perform cVNTs is lacking.

Keywords: Luminex; Neutralizing antibody test; SARS-CoV-2; Serosurveillance; cPass.

Fig. 1

Correlations between the percentage of inhibition measured by the surrogate viral neutralisation test (sVNT) and the log(dilution factors or signal-to-noise ratio) for the conventional viral neutralization test (cVNT) or the Luminex multiplex immunological assay (MIA) as calculated by the nonparametric Spearman correlation test (r_s). Seropositivity cut-off levels for the sVNT are indicated by the dashed grey lines at 20 or 30 % inhibition. Negative samples on the cVNT or MIA were not included in these figures.