(P-PH-2) Investigating the relevance of anti-spike nonreactive, anti-nucleocapsid reactive profiles identified in a large SARS-CoV-2 blood donor serosurveillance study

Background/Case Studies: Following SARS-CoV-2 infection most individuals develop anti-spike (S) and anti-nucleocapsid (N) responses. To estimate SARS-CoV-2 donor incidence, routine donations were tested using the Ortho VITROS S Total Ig assay followed by testing S-reactive donations (S+) by the Roche Elecsys N Total Ig assay. To streamline workflow, in Sep 2021 testing transitioned to simultaneous S+N testing on the VITROS platform and the Ortho N assay, with slightly enhanced sensitivity, replaced the Roche N assay. Parallel testing enabled the identification of donations with S-nonreactive but N-reactive (S-/N+) results.

Study

Design/Methods: To further characterize the S-/N+ serologic response for consistency, serum samples from S-/N+ donations were reassessed by repeat testing using the Ortho N assay, and further evaluated by the Roche N and the V-PLEX SARS-CoV-2 panel 2 IgG (MSD) assays. Donations that were reactive by Ortho N and an independent serologic assay (Roche N or MSD) were considered presumptive positive (PP); those concordant-reactive by all assays were classified as confirmed positive (CP).
To gain insight on the donor infection status (i.e., current vs past infection) donations were assessed for the presence of viral RNA using the Grifols Procleix SARS-CoV-2 transcription-mediated amplification (TMA) assay. Donor responses to a survey on SARS-CoV-2 infections and symptoms were reviewed.

Results/Findings: Of routine donations screened between Sep 2021 – Feb 2022, 0.1% (487/469,494) were S-/N+. Most donations tested either S+/N+ (23%; 109,914/469,494) or S+/N- (71%; 331,443/469,494) during the same period. The average Ortho N signal to cut-off ratio (S/CO) for the S-/N+ donations was 9.12. Of 159 S-/N+ donations available for further testing, 57% (90/159) were repeat reactive (RR) when retested with the Ortho N assay, with an average S/CO of 12.34. Retesting by independent N-assays identified 50 PP (31%) and 24 (15%) CP donations. All 159 donation samples were SARS-CoV-2 RNA (TMA) nonreactive suggesting that the RR, PP and CPs corresponded to past infections; the remainder were likely false positives (FP). Of the 34 donors responding to the survey, 18 (53%) also reported a history of infection.

Conclusions: Approximately half of S-/N+ donations (~0.05%) were likely FP; a rate substantially less than that reported by the manufacturer of 0.9%. The 15-57% of retested S-/N+ donations (or 0.005-0.019% of 469,494 screened donations) testing reactive by multiple N assays in the absence of viral RNA and S antibody reactivity likely represent very low-level reactivity resulting from remote infections, thus S-/N+ donations have little to no impact on seroprevalence estimates.

Importance of research: Changes to blood donation antibody screening algorithms and testing platforms in national SARS-CoV-2 serosurveillance programs can result in the identification of low frequency serological profiles. Donation samples testing nucleocapsid antibody reactive in the absence of spike antibody and viral RNA reactivity likely represent either false positive results or remote infections with little to no impact on incidence estimates of infection.