Qualification of COVID-19 convalescent plasma donations by SARS-CoV-2 antibody testing utilizing a wide-dynamic range Antigen-Sandwich S1-antigen immunoassay and reflex neutralizing antibody testing

ABSTRACT

Background/Case Studies The efficacy of COVID-19 convalescent plasma (CCP) to treat COVID-19 is hypothesized to be associated with the concentration of neutralizing antibodies (nAb) to SARS-CoV-2. While high capacity, automated serologic assays to detect binding antibodies (bAb) have been developed, complex nAb assays are not easily adaptable to high-throughput testing. We sought to determine the effectiveness of using surrogate bAb signalto- cutoff ratio (S/CO) in predicting nAb titers using a pseudovirus reporter viral neutralization (RVPN) assay. Study Design/Methods: CCP donor serum collected by 3 large US blood collectors was tested with a bAb assay (Ortho Diagnostic VITROS® Anti-SARS-CoV-2 Total, CoV2T) and a nAb RVPN assay. Although EUA approved as a qualitative assay, CoV2T reports a semi-quantitative S/CO. The RVPN assay uses a pseudovirus construct with native S-protein and target cell lines overexpressing ACE2 receptor and TMPRSS2 protease. Serially diluted serum is mixed with SARS-CoV-2 pseudovirus to assess inhibition of viral entry in culture and reported as titers resulting in 50% neutralization of virus infectivity (NT50) by nonlinear regression analysis. CoV2T prediction effectiveness at several S/CO thresholds was evaluated for various RVPN nAb NT50 titers using receiver operating characteristic analysis. Results/Findings: 753 CCP donations were tested with median CoV2T S/CO of 71.2 (range 0.1-919) and median NT50 of 527.5 (range <40 to >10,240). The prevalence of CCP donors with NT50 over various target n-Ab titers were 86% >80, 76% >160, and 45% >640. Increasing CoV2T reactivity threshold reduces sensitivity to predict the target NT50 titer while specificity to identify those below nAb threshold increases for all targeted NT50s (Table 1). As the targeted NT50 is increased from >80 to >640, the positive predictive value falls dramatically while the negative predictive value increases, thus S/CO thresholds are less able to predict donors who have the target NT50 titer but more able to predict those donors who do not meet it. Conclusions: The selection of targeted nAb titer for clinical use will significantly impact availability of CCP for transfusion. Product release with CoV2T assay S/CO thresholds must balance the risk of releasing products below minimum target nAb titer and the cost of false negatives (CCP units below the threshold with adequate nAb titers). A two-step testing scheme may be optimal, with nAb testing performed on CoV2T reactive samples with S/CO values below the release threshold.