MSD V-PLEX SARS-CoV-2 ASSAY to IDENTIFY VACCINATED and INFECTED INDIVIDUALS

ABSTRACT

Background: The prevalence of vaccinated, previously infected, and individuals at risk of SARS-CoV-2 infection is important for epidemiologic studies and public health interventions. Asymptomatic infections and reluctance to disclose vaccination status hinder accurate assessments of the current state of the epidemic. Since COVID-19 vaccines generate immune responses to spike (S1), but not nucleocapsid (N), it is possible to differentiate between vaccinated, infected, and unexposed individuals by comparing antibody reactivity to each antigen. The MSD V-Plex SARS-CoV-2 IgG assay can potentially differentiate each state in one test by simultaneously evaluating IgG reactivity to the S1, receptor binding domain (RBD), and N proteins. Methods: The MSD assay was validated with three sample sets known vaccination with no previous infection (n=158);known infected and not vaccinated (n=157);and samples collected prior to the COVID-19 pandemic in 2016 (n=144). Of the previously infected individuals, 15 (9.6%) were hospitalized;sample collection occurred a median of 48 days after a PCR-positive result. Using an algorithm, samples with positive results on both S1 and RBD but negative on N were classified as vaccinated. Samples with a positive result on all three proteins were considered to be infected with the possibility of subsequent vaccination. Any other result was classified as unexposed. Sensitivity and specificity for each state were calculated. Results: Reactivity to each antigen is shown in the figure. 100% (95% confidence interval [CI] 97.7-100), 92% (95% CI 86.3-95.5), and 0.7% (95% CI 0.02-3.8) of vaccinated, infected, and unexposed samples were positive for S1. 100% (95% CI 97.7-100.0%), 91% (95% CI 85.5-95.0%), and 0.7% (95% CI 0.02-3.8%) of vaccinated, infected and unexposed samples were positive for RBD. 0% (95% CI 0-2.3), 86% (95% CI 79.6-91.0), and 2.1% (95% CI 0.4-6.0) of vaccinated, infected and unexposed samples were positive for N. Algorithm sensitivity and specificity for classification of vaccinated samples were 100% (95% CI 97.7-100) and 96.7 (95% CI 94-98.4). For the classification of samples from previously infected individuals, sensitivity and specificity were 83.4% (95% CI 76.7-88.9) and 100% (95% CI 98.8-100). Conclusion: This study establishes the sensitivity and specificity for a high-throughput assay ideal for SARS-CoV-2 seroprevalence studies. Future research should focus on applying this assay in health care settings to guide practice and policy to mitigate the pandemic.