Evaluation of the performance of a multiplexed serological assay in the detection of SARS-CoV-2 infections in a predominantly vaccinated population

Abstract

SARS-CoV-2 seroprevalence studies may be complicated by vaccination efforts. It is important to characterize the ability of serology methods to correctly distinguish prior infection from postvaccination seroreactivity. We report the performance of the Meso Scale Discovery (MSD) V-PLEX COVID-19 Coronavirus Panel 2 IgG assay. Using serum samples from a prospective cohort of paramedics, we calculated the performance of the V-PLEX nucleocapsid (“N”) assay to classify prior SARS-CoV-2 infections, defined as a (i) history of a positive SARS-CoV-2 PCR test or (ii) positive serology results using the Roche Elecsys total nucleocapsid anti-SARS-Cov-2 assay. We calculated sensitivity and specificity at the optimal threshold (defined by the highest Youden index). We compared subgroups based on vaccination status, and between models that excluded prior infections 3 to 12 months before sample collection. Of 1119 participants, 914 (81.7%) were vaccinated and 60 (5.4%) had evidence of a preceding SARS-CoV-2 infection. Overall and within vaccinated and unvaccinated subgroups, the optimal thresholds were 828 AU/mL, 827 AU/mL, and 1324 AU/mL; with sensitivities of 0.95 (95% CI: 0.94 to 0.96), 0.95 (0.94 to 0.96), 0.94 (0.92 to 0.96) and specificities of 0.88 (0.86 to 0.90), 0.87 (0.85 to 0.89), and 0.94 (0.89 to 0.98), respectively. N-assay specificity was significantly better in unvaccinated (versus vaccinated) individuals (P = 0.005). Overall optimal thresholds based on the AUC values were higher for samples from unvaccinated participants, especially when examining infections within the preceding 9 months (5855 versus 1704 AU/mL). Overall, V-PLEX nucleocapsid assay cutoff values were higher among unvaccinated individuals. Specificity was also significantly higher among unvaccinated individuals. Different thresholds were required to achieve optimal test performance, especially for detecting SARS-CoV-2 infections within the preceding 9 months.

IMPORTANCE Among a cohort of adult paramedics in Canada, we investigated the performance of nucleocapsid (N) antibody detection (measured with a V-PLEX assay) to identify previous COVID-19 infections and compared differences among vaccinated and unvaccinated. Our data indicate that vaccinated and unvaccinated groups require different thresholds to achieve optimal test performance, especially for detecting COVID-19 within the preceding 9 months. Overall, specificity was significantly higher among unvaccinated, compared to vaccinated individuals.