SARS‐CoV‐2 antibody persistence in COVID‐19 convalescent plasma donors: Dependency on assay format and applicability to serosurveillance

Di Germanio, Clara; Simmons, Graham; Kelly, Kathleen; Martinelli, Rachel; Darst, Orsolya; Azimpouran, Mahzad; Stone, Mars; Hazegh, Kelsey; Grebe, Eduard; Zhang, Shuting; Ma, Peijun; Orzechowski, Marek; Gomez, James E.; Livny, Jonathan; Hung, Deborah T.; Vassallo, Ralph; Busch, Michael P.; Dumont, Larry J.

BACKGROUND: Antibody response duration following severe acute respiratory syndrome coronavirus 2 infection tends to be variable and depends on severity of disease and method of detection. STUDY DESIGN AND METHODS: COVID‐19 convalescent plasma from 18 donors was collected longitudinally for a maximum of 63–129 days following resolution of symptoms. All the samples were initially screened by the Ortho total Ig test to confirm positivity and subsequently tested with seven additional direct sandwich or indirect binding assays (Ortho, Roche, Abbott, Broad Institute) directed against a variety of antigen targets (S1, receptor binding domain, and nucleocapsid [NC]), along with two neutralization assays (Broad Institute live virus PRNT and Vitalant Research Institute [VRI] Pseudovirus reporter viral particle neutralization [RVPN]). RESULTS: The direct detection assays (Ortho total Ig total and Roche total Ig) showed increasing levels of antibodies over the time period, in contrast to the indirect IgG assays that showed a decline. Neutralization assays also demonstrated declining responses; the VRI RVPN pseudovirus had a greater rate of decline than the Broad PRNT live virus assay. DISCUSSION: These data show that in addition to variable individual responses and associations with disease severity, the detection assay chosen contributes to the heterogeneous results in antibody stability over time. Depending on the scope of the research, one assay may be preferable over another. For serosurveillance studies, direct, double Ag‐sandwich assays appear to be the best choice due to their stability; in particular, algorithms that include both S1‐ and NC‐based assays can help reduce the rate of false‐positivity and discriminate between natural infection and vaccine‐derived seroreactivity.